UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects of the solid tumour microenvironment on the penetration and distribution of trastuzumab Sy, Jonathan Tiu 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2010_fall_sy_jonathan.pdf [ 11.85MB ]
JSON: 24-1.0071376.json
JSON-LD: 24-1.0071376-ld.json
RDF/XML (Pretty): 24-1.0071376-rdf.xml
RDF/JSON: 24-1.0071376-rdf.json
Turtle: 24-1.0071376-turtle.txt
N-Triples: 24-1.0071376-rdf-ntriples.txt
Original Record: 24-1.0071376-source.json
Full Text

Full Text

   EFFECTS OF THE SOLID TUMOUR MICROENVIRONMENT ON THE PENETRATION AND DISTRIBUTION OF TRASTUZUMAB   by   Jonathan Tiu Sy  B.Sc., University of British Columbia, 2003     A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE   in   The Faculty of Graduate Studies  (Pathology and Laboratory Medicine)    THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   October 2010    © Jonathan Tiu Sy, 2010     ii ABSTRACT   The abnormal development of the solid tumour microenvironment may result in populations of tumour cells that receive sub-optimal drug exposure and as a consequence experience diminished efficacy of treatment.  Antibody-based therapies may face additional limitations from their inherent molecular size and antigen binding specificity.  Trastuzumab, a monoclonal antibody that targets the HER2 receptor, has been shown to have incomplete distribution in HER2 over- expressing tumour xenografts.  The objective of this research was to examine the roles of various tumour microenvironment components, namely pericytes and epithelial tight junctions, in modulating extravasation of trastuzumab and its distribution throughout the tumour.  To examine pericyte coverage, tumour bearing mice were administered imatinib to reduce pericyte recruitment. Fluorescent microscopy was used to assess blood vessel changes and directly visualized the effects these changes had on systemically administered trastuzumab.  Imatinib treatment resulted in increased necrosis, decreased microvessel density and pericyte staining, and a reduction in the fraction of blood vessels in close proximity to pericytes.  Interestingly, trastuzumab distribution became significantly reduced in imatinib treated tumours and blood vessel analysis revealed vessels possessing pericyte coverage to have considerably greater trastuzumab penetration. These results indicate that combinations of imatinib prior to trastuzumab dosing may not be optimal for purposes of increased drug distribution.  However, the degree of pericyte coverage may contribute to the heterogeneous distribution of trastuzumab.  To investigate the  iii role of epithelial tight junctions (TJ), MCF7HER2 multilayered cell cultures (MCC) were exposed to trastuzumab and bound antibody was visualized by immunohistochemistry. Trastuzumab was found to penetrate effectively from the “basolateral” side of the MCC attached to the anchoring membrane, whereas penetration into the apical, or “luminal,” cell layer was limited to the outermost row of cells even after prolonged exposure.  Transmission EM revealed confluent TJ in the luminal cell layer and immunohistochemical staining of the peripheral TJ marker ZO-1 revealed positive staining in a continuous band along the luminal cell layer that was consistent with the localization of bound trastuzumab. Taken together, these data suggest that abnormal pericyte coverage and TJ localization are obstacles in the tumour microenvironment preventing the effective distribution of anti-cancer agents such as trastuzumab.   iv PREFACE The preparation of electron microscope samples following gluteraldehyde fixation (detailed in section 2.9) was carried out by Fanny Chu.  Handling of the electron microscope for image acquisition was performed by Dr. David Walker. Both procedures took place at the St. Paul’s Hospital iCapture Centre.  I was responsible for preparation and fixation of my cell samples and image selection. All animal protocols were approved by the University of British Columbia Animal Care Committee, certificate #A07-0404 (see Appendix), and all studies were performed according to Canadian Council of Animal Care guidelines.  v TABLE OF CONTENTS   Abstract .............................................................................................................  ii  Preface ............................................................................................................... iv  Table of Contents .............................................................................................  v  List of Figures .................................................................................................  vii  Abbreviations ...................................................................................................  ix  Acknowledgements .........................................................................................  xi  Dedication .......................................................................................................  xii  1 Introduction ..................................................................................................  1  1.1  Breast cancer ......................................................................................  1   1.1.1 HER2 overexpressing breast cancer ........................................  3   1.1.2 HER2 inhibition by trastuzumab................................................  4  1.2  Solid tumour microenvironment...........................................................  6   1.2.1 Blood vessel architecture..........................................................  6   1.2.2 Vascular endothelial cells .........................................................  7   1.2.3 Pericytes...................................................................................  9   1.2.4 Tight junctions.........................................................................  11  1.3  Hypothesis and objectives.................................................................  14  2 Materials and Methods ..............................................................................  20  2.1  Monolayer cell culture........................................................................  20  2.2  Verification of surface HER2 expression by flow cytometry...............  20  2.3  Tumour xenograft model ...................................................................  21  2.4  Immunohistochemistry.......................................................................  23  2.5  Image acquisition ..............................................................................  24  2.6  Image analysis ..................................................................................  24  2.7  Multilayered cell culture.....................................................................  25  2.8  MCC drug treatment and tissue collection.........................................  26  2.9  Transmission electron microscopy ....................................................  27  2.10 Statistical analysis .............................................................................  27  3 Investigating the Role of Pericytes in Trastuzumab Distribution .........  30  3.1  Rationale and hypothesis .................................................................  30  3.2  Results ..............................................................................................  33   3.2.1 Extravascular distribution of trastuzumab over time ...............  33   3.2.2 Trastuzumab downregulates HER2 surface expression.........  34  vi   3.2.3 Trastuzumab induces a growth delay in MFC7HER2 tumours ..  35   3.2.4 Distribution of different MW dextrans relative to vasculature     is size-dependent....................................................................  36   3.2.5 Binding of HER2 by trastuzumab increases dextran     delivery to tumours .................................................................  36   3.2.6 Increased necrosis, decreased MVD, and decreased     pericyte coverage following imatinib treatment .......................  37   3.2.7 Trastuzumab is decreased in imatinib treated tumours ..........  38   3.2.8 Extravascular penetration of trastuzumab is increased     in blood vessels positive for pericyte coverage.......................  39  3.3  Discussion .........................................................................................  39  4 Investigating the Role of Tight Junctions in Trastuzumab  Distribution.................................................................................................  59  4.1  Rationale and hypothesis .................................................................  59  4.2  Results ..............................................................................................  62   4.2.1 Luminal and basolateral sides of MCF7HER2 MCC     exhibit dissimilar trastuzumab penetration..............................  62   4.2.2 Luminal tight junction localization in transmission electron     micrographs............................................................................  62   4.2.3 Tight junctions restrict trastuzumab penetration in MCC ........  63   4.2.4 Tight junctions form discontinuous and haphazard     patterns in MCF7HER2 tumour xenografts ................................  64  4.3  Discussion .........................................................................................  65  5 General Conclusions ................................................................................  79  5.1  Summary and future directions..........................................................  79  References.......................................................................................................  82  Appendix: Animal Care Certificate.................................................................  93   vii LIST OF FIGURES   Figure 1.1 Ductal carcinoma formation in breast...........................................  17  Figure 1.2 Pericytes and endothelial cells ....................................................  18  Figure 1.3 Junctional complexes ..................................................................  19  Figure 2.1 Multilayered cell culture growth chamber ....................................  29  Figure 3.1 Immunohistochemical imaging of bound trastuzumab in  trastuzumab treated MCF7HER2 tumour xenografts ......................  48  Figure 3.2 HER2 expression in MCF7HER2 xenografts following  Treatment with 4 mg/kg trastuzumab...........................................  49  Figure 3.3 Growth delay of MCF7HER2 xenografts following single  or multiple doses of 4 mg/kg trastuzumab ...................................  50  Figure 3.4 Immunohistochemical imaging of different molecular weight  FITC-dextrans in MCF7HER2 xenografts........................................  51  Figure 3.5 Increased dextran delivery following 4 mg/kg trastuzumab  treatment .....................................................................................  52  Figure 3.6 FITC-Dextran controls .................................................................  53  Figure 3.7 Increased blood vessel permeability following 4 mg/kg  trastuzumab treatment ................................................................  54  Figure 3.8 Effects of imatinib mesylate on the vasculature of MCF7HER2  xenografts ...................................................................................  55  Figure 3.9 Imatinib treatment decreases trastuzumab distribution in  MCF7HER2 xenografts ..................................................................  56  Figure 3.10 Increased extravascular penetration of trastuzumab from  blood vessels positive for pericyte coverage ...............................  58  Figure 4.1 MCF7HER2 MCC exposed to tratuzumab from the luminal  and basolateral sides of the culture .............................................  73  Figure 4.2 Trastuzumab penetration profile in MCF7HER2 MCC ....................  74   viii Figure 4.3 Transmission electron micrographs of MCF7HER2 MCC  junctional complexes ...................................................................  75  Figure 4.4 Immunohistochemical imaging of ZO-1 positive tight junctions ...  76  Figure 4.5 Tight junctions on the luminal surface of MCC restrict  trastuzumab penetration ..............................................................  77  Figure 4.6 Tight junctions in MCF7HER2 xenografts ......................................  78   ix ABBREVIATIONS   Abbreviation Definition  α-SMA Alpha-smooth muscle actin ADCC Antibody-dependent cell-mediated cytotoxity ARC Animal Research Centre BrdUrd Bromodeoxyuridine CD105 Cluster of differentiation 105 CD31 Cluster of differentiation 31 CDK2 Cyclin-dependent kinase 2 CDR Complementarity determining region Da Dalton DCIS Ductal carcinoma in situ DiOC7 3,3' - Diheptyloxacarbocyanine iodide ECM Extracellular matrix EDTA Ethylenediaminetetraacetate EGF Epidermal growth factor EGFR Epidermal growth factor receptor ER Estrogen receptor FDA Food and Drug Administration FISH Fluorescence in situ hybridization FITC Fluorescein isothiocyanate HER2 Human epidermal growth factor receptor 2 HGF Human growth factor HIF-1α Hypoxia inducible factor-1 alpha HRP Horseradish peroxidase IDC Invasive ductal carcinoma IgG Immunoglobulin G IHC Immunohistochemistry IP Intraperitoneal IV Intravenous  x JAM Junction adhesion molecule kb Kilo-base pairs MAGUK Membrane-associated guanylate kinase MBC Metastatic breast cancer MCC Multilayered cell culture MDCK Madin-Darby Canine Kidney MEM Minimum essential medium MESF Molecules of soluble fraction MRFI Mean relative fluorescence intensity MRI Magnetic resonance imaginc mRNA Messenger ribonucleic acid MVD Microvessel density NOD/SCID Non-obese diabetic / severe combined immunodeficiency NSCLC Non-small cell lung cancer PDGF-β Platelet-derived growth factor-beta PDGFR-β Platelet-derived growth factor receptor-beta PgR Progesterone receptor PI3K/AKT Phosphatidylinositol 3-kinases RPA Reverse passive Arthus RPMI Roswell Park Memorial Institute SD Standard deviation TEER Transepithelial electrical resistance TEM Transmission electorn microscopy TER Transepithelial resistance TGF-β Transforming growth factor-beta TJ Tight junction VEGF Vascular endothelial growth factor VEGFR2 Vascular endothelial growth factor receptor 2 ZO-1 Zonula occludin-1   xi ACKNOWLEDGEMENTS   First, I would like to thank Drs. Wieslawa Dragowska and Donald Yapp for their help with the cell lines and surface expression assays.  I would also like to thank Dr. David Walker and the iCapture Centre for generously providing their time and guidance in the electron microscopy experiments.   I want to give a special thanks to AIM lab members Drs. Jennifer Baker and Alastair Kyle.   Jennifer who provided much insight and technical assistance for my immunohistochemistry and animal work, and Alastair who contributed countless ideas and taught me all the laboratory techniques necessary to complete the MCC studies.   I would also like to thank my graduate committee, Drs. Mladen Korbelik, Pamela Hoodless, and Wan Lam, for their optimism and ongoing encouragement through the ups and downs of this project.   Lastly, I would like to give my utmost thanks to my supervisor Dr. Andrew Minchinton, who provided me with time and support to conduct my research in the direction I was most interested in pursuing.  His openness and constructive feedback was an inspiration during the course of my research.  xii DEDICATION   This work is dedicated to my loving parents who gave me endless encouragement and support to see this work through.     1 1 INTRODUCTION  1.1 Breast cancer  In women, breast cancer is the most common form of cancer (Parkin, Bray, Ferlay, & Pisani, 2005) and the leading cause of cancer death worldwide (World Health Organization, 2009).  Breast cancer survival rates vary greatly worldwide, with 5-year rates ranging from 85% in developed countries to 50-60% in developing countries (International Agency for Research on Cancer, 2008). Lower survival rates in less developed countries are attributed mainly to a lack of early detection programs, resulting in a higher proportion of women presenting with late-stage disease (Coleman, et al., 2008).   Most breast cancers originate from either the inner epithelial linings of the milk-supplying lobules or ducts of the breast (Figure 1.1).  Of these, over 80% originate in the ductal epithelium while the remainder originate in cells of the lobular epithelium (International Agency for Research on Cancer, 2008).  In ductal carcinomas, prior to malignancy, the cells of the ductal lining will undergo abnormal growth into phenotypically heterogeneous lesions, termed ductal carcinoma in situ (DCIS)  (Maass, Alkasi, Bauer, Jonat, Souchon, & Meinhold- Heerlein, 2009).  Although DCIS is a non-obligate precursor of an invasive phenotype, it is proposed that cellular events such as oncogene activation and stromal modifications can advance DCIS to a malignant state, enabling the cancer cells to break through the basement membrane and invade the  2 surrounding stromal tissue as an invasive ductal carcinoma (IDC)  (Damonte, Hodgson, Chen, Young, Cardiff, & Borowsky, 2008).  Metastatic breast cancer (MBC) is the most advanced stage of breast cancer and is currently uncurable. Metastasis describes cancer cells at the original site that have intravasated the nearby vasculature, circulated through the bloodstream, and established secondary tumours at distant organs of the body.  In MBC, these are most commonly to the bones, lungs, and liver  (Tait, Waterworth, Loncaster, Horgan, & Dodwell, 2005).   Treatment of breast cancer depends on the stage of the tumour, as classified by the tumour size, axillary lymph node involvement, and presence of metastases, level of differentiation, and also by the receptor status of the tumour cells  (Escobar, Patrick, Rybicki, Weng, & Crowe, 2007).  Of the latter, breast cancers are categorized by the presence of the estrogen receptor (ER), progesterone receptor (PgR), and the human epidermal growth factor receptor 2 (HER-2).  Nearly 70% of breast cancers are positive for the estrogen receptor (ER) and/or progesterone receptor (PgR)  (Pritchard, 2005).  Breast cancers that are positive for the ER or PgR, an ER-dependent gene product, have a dependence on the estrogen hormone for development and growth  (Massarweh & Schiff, 2006).  Consequently, patients with ER and/or PR positive cancer cells are administered endocrine treatments that target these receptors, such as agents that block ER-ligand interaction (selective estrogen receptor modulators) or reduce estrogen production (aromatase inhibitors)  (Hiscox, Davies, & Barrett-  3 Lee, 2009).  Similarly, treatments including trastuzumab, an antibody-based molecular targeting approach, is used to treat patients with HER-2 overexpressing breast cancer.  1.1.1 HER-2 overexpressing breast cancer  HER-2/neu is a proto-oncogene that encodes the transmembrane tyrosine kinase receptor protein HER-2 (also known as ERBB-2).  HER-2 is a member of closely related growth factor receptors that includes the epidermal growth factor receptor (EGFR/HER-1), HER-3, and HER-4  (Ross & Fletcher, 1998).  All members contain an extracellular ligand-binding domain, a transmembrane region, and a cytoplasmic tyrosine-kinase domain.  Binding of EGFR, HER-3, and HER-4 to their epidermal growth factor (EGF) ligands induces receptor homo- and herterodimer formation and activation of the kinase domain, leading to subsequent phosphorylation and activation of downstream signaling pathways involved in cell growth, differentiation, and survival  (Yarden & Sliwkowski, 2001). Although HER-2 has no known EGF ligand, it is important because it is the preferred heterodimerization partner of the other ligand-binding family members (Graus-Porta, Beerli, Daly, & Hynes, 1997).   Overexpression of the HER-2 oncogene occurs in approximately 20% of breast cancers and generally arises due to gene amplification  (Wolff, et al., 2007).  HER-2 overexpression results in 50-100 fold increase in the number of HER-2 surface receptors compared to normal mammary epithelial cells  (Wilson,  4 et al., 2005).  Verification of the HER-2 receptor status is performed using either fluorescence in-situ hybridization (FISH) to detect the gene amplification or by immunohistochemical analysis to confirm overexpression of the HER-2 protein.   In breast cancer patients, HER-2 overexpression has been correlated with increased disease recurrence and poor prognosis  (Slamon, Clark, Wong, Levin, Ullrich, & McGuire, 1987).  EGF family activation and signaling has been shown to increase expression of proteins involved in cell proliferation, apoptosis, and cell cycle regulation through pathways such as the phosphatidylinositol-3-kinase (PI3K)/AKT pathway  (Hennessy, Smith, Ram, Lu, & Mills, 2005).  HER-2 overexpression has been found to increase vascular endothelial growth factor (VEGF) expression, independent of hypoxic conditions, as well as increasing hypoxia inducible factor-1 α (HIF-1α) synthesis  (Petit, et al., 1997)  (Laughner, Taghavi, Chiles, Mahon, & Semenza, 2001).  HER-2 positive cells have also shown reduced response to the anti-proliferative transforming growth factor-β (TGF-β), which is involved in cell cycle and apoptosis regulation  (Wilson, et al., 2005).  1.1.2 HER2 inhibition by trastuzumab  Trastuzumab (Herceptin, Genentech) is a recombinant humanized monoclonal antibody used to treat breast cancer patients that overexpress HER- 2.  Humanization of trastuzumab refers to the engineering of the complementarity determining region (CDR) of a mouse antibody targeting the extracellular domain  5 of HER-2 onto a human antibody fragment, thus allowing specificity for the human HER-2 receptor while minimizing an adverse host immune response (Booy, et al., 2006).  Patients treated with trastuzumab experience significantly slower tumor growth, regression in tumour size, and improved disease-free and overall survival rates in HER2-overexpressing breast cancer  (Romond, et al., 2005).  Trastuzumab was approved by the Food and Drug Administration (FDA) in 1998 as a first-line treatment for metastatic breast cancer and in 2006 in combination with chemotherapy drugs doxorubicin, cyclophosphamide, and paclitaxel for early stage breast cancer adjuvant therapy  (Lin & Rugo, 2007).   Trastuzumab is administered systemically to patients and functions primarily by accessing the surface of breast cancer cells via the tumour vasculature and binding the extracellular domain of the HER-2 receptor.  The precise mechanisms by which trastuzumab acts on the HER-2 overexpressing cells is still not completely defined, but several molecular and cellular effects have been observed.  Exposing in vitro cell cultures to trastuzumab resulted in internalization and degradation of the HER-2 receptor  (Cuello, et al., 2001). Trastuzumab binding also disrupted HER-2 activation of the PI3K/AKT signaling pathway  (Yakes, Chinratanalab, Ritter, King, Seelig, & Arteaga, 2002) and decreased the activity of cell cycle protein CDK2  (Lane, Motoyama, Beuvink, & Hynes, 2001).  In a breast cancer xenograft mouse model, trastuzumab treatment reduced levels of VEGF and decreased microvessel density in vivo (Izumi, Xu, di Tomaso, Fukumura, & Jain, 2002).  There is also evidence of  6 immune effects by antibody dependent cell-mediated cytotoxicity (ADCC). Strong lymphoid infiltration of tumours was noted in patients treated with pre- operative trastuzumab  (Gennari, et al., 2004) and natural killer cell-mediated ADCC has been observed in multiple breast cancer cell lines  (Clynes, Towers, Presta, & Ravetch, 2000) (Cooley, Burns, Repka, & Miller, 1999).  1.2 Solid tumour microenvironment  The term “tumour microenvironment” is used to describe the non-cancer cell components that make up the solid tumour stroma.  These include vascular endothelial cells that form blood vessels, the connective tissue of the extracellular matrix (ECM) and its various molecular components, as well as pericytes, fibroblasts, inflammatory cells, and leukocytes of the immune system. As a cancer progresses, the surrounding microenvironment becomes co-opted by continuous paracrine signaling of the cancer cells, creating irregularities in stromal morphology and function that promote tumour initiation and progression (Pietras & Ostman, 2010).  Targeted therapies focused on modulating components of the tumour microenvironment will provide alternative strategies to the conventional targeting of cancer cells.  1.2.1 Blood vessel architecture  Solid tumour growth and metastasis is dependent on the recruitment of new blood vessels for delivery of oxygen and nutrients to the growing mass of cancer cells.  Angiogenesis, the process of formation of new blood vessels from  7 pre-existing vessels or endothelial cell progenitors, is governed by the tumour’s production of diffusible pro-angiogenic and anti-angiogenic factors  (Carmeliet & Jain, 2000).  The tumor microenvironment contains excessive amounts of pro- angiogenic factors derived from neoplastic, stromal, and infiltrating immune cells (Hall, 2006). The imbalance of pro-angiogenic and anti-angiogenic factors promotes abnormal angiogenesis, creating numerous blood vessels with structural abnormalities and functional defects.   Although pro-angiogenic factors are favored in solid tumours, the homeostatic regulation between cancer cell growth and development of blood vessels is impaired, and tumour cells will often proliferate more rapidly than the cells that form the vasculature  (Denekamp & Hobson, 1982).  The lagging vascularization of the growing tumour results in populations of cancer cells that exist at relatively large distances from blood vessels (>100 µm), giving rise to the cancer hallmark hypoxia and imposing an obstacle to the delivery of anti-cancer drugs to these regions of the tumour  (Minchinton & Tannock, 2006).  In addition, the architecture of the tumour vasculature is highly disorganized, with tortuous and dilated vessels exhibiting uneven diameters, excessive branching, and shunts  (Carmeliet & Jain, 2000), exacerbating the chaotic structure of solid tumours.   8 1.2.2 Vascular endothelial cells  The extensive heterogeneity of tumour vessels is also observed at the level of the endothelial cell lining, whereby some vessels are aberrantly formed compared to the monolayers seen in normal tissues.  Such endothelial cells have been characterized as being disorganized, irregularly shaped, and having loose interconnections and focal intercellular openings  (McDonald & Foss, 2000).  The abnormal structure of these vessels provides endothelial fenestrations, transcellular holes, widened inter-endothelial junctions, and a discontinuous or absent basement membrane, all of which contribute to the high vascular permeability, or “leakiness,” of tumour blood vessels  (Hashizume, et al., 2000). However, not all vessels within a tumour are leaky.  Vascular permeability varies not only from one tumour to the next, but it also varies spatially and temporally within the same tumour  (Fukumura & Jain, 2007).   Erratic blood vessel structure and leakiness, in combination with haphazard vascular patterns of the tumour vasculature, is responsible for the irregular and intermittent blood flow that has been observed in solid tumours (Toffoli & Michiels, 2008).  Lymphatic drainage of fluids is also impaired in tumours, with reduced or absent lymphatics found within the central tumour region  (Leu, Berk, Lymboussaki, Alitalo, & Jain, 2000).  One explanation may be the rapid growth of surrounding tumour tissue that leads to compression of both blood vessels and lymphatic channels  (Helmlinger, Netti, Lichtenbeld, Melder, & Jain, 1997).  Lack of an efficient lymphatic system and reduced blood flow are  9 contributors to the occurrence of elevated interstitial fluid pressure (IFP) in solid tumours, which may interfere with convective extravascular delivery of therapeutic agents  (Flessner, Choi, Credit, Deverkadra, & Henderson, 2005).  1.2.3 Pericytes  Blood vessels are composed of the endothelial cells that form the inner lining of the vessel wall and mural cells that surround the endothelial tubing (Figure 1.2).  These mural cells, called vascular smooth muscle cells when encircling larger vessels, are referred to as pericytes when they reside in the wall of smaller vessels such as capillaries.  Pericytes possess long cytoplasmic processes that can contact multiple endothelial cells and serve to communicate information along the vessel both physically and through paracrine signaling (Bergers & Song, 2005). These contacts are composed of peg-and-socket junctions, adhesion plaques, and gap junctions, which enable the cells to connect with each other through discontinuities in the basement membrane  (Rucker, Wynder, & Thomas, 2000).  Functionally, pericytes are mainly associated with the stabilization and hemodynamic processes of blood vessels.   Similar to the smooth muscle of larger arteries, pericytes regulate capillary blood flow through contractile transmission in response to surrounding hemodynamic forces and vasoactive signals  (Rucker, Wynder, & Thomas, 2000).  Pericyte and endothelial cell interaction has also been shown to play an important role in angiogenesis and vascular development.  In a common model  10 of vessel formation, angiogenic factors such as VEGF stimulate endothelial cells to secrete proteases to degrade the vessel basement membrane.  This allows the endothelial cells to invade the ECM and form a migration column that moves toward the VEGF gradient  (Gerhardt & Betsholtz, 2003).  The newly formed sprouts continue to extend while secreting growth factors, particularly platelet- derived growth factor β (PDGF-β), to attract pericytes that possess the PDGF-β receptor (PDGFR-β) and promote vessel maturation  (Betsholtz, Lindblom, & Gerhardt, 2005).  Recruited pericytes provide vessel stability and transmit angiogenic signals along the endothelial cell column that induce endothelial differentiation and growth arrest  (Hall, 2006).  Because the recruitment of pericytes lags behind endothelial sprouting, the timing of pericyte coverage can result in endothelial growth, regression, or stabilization depending on the presence or absence of angiogenic growth factors.   In contrast to normal tissue blood vessels, tumour pericytes are visibly reduced in density and appear more loosely associated with the vasculature. This was demonstrated in multiple tumour types, including pancreatic islet carcinoma, glioblastoma, and mammary carcinoma, using transgenic or orthotopic mouse models  (Bergers & Song, 2005).  These pericytes were observed to have cytoplasmic processes that were aberrant, extending away from the endothelial cells and into the tumour tissue.  There was also variability in the degree of pericyte coverage between tumour types, with pancreatic islet carcinomas having relatively dense coverage and glioblastomas and mammary  11 carcinomas having a reduced pericyte density.  Additionally, the pericyte expression profile of common molecular markers, particularly α-smooth muscle actin (α-SMA), has been seen to vary between tumour types  (Morikawa, Baluk, Kaidoh, Haskell, Jain, & McDonald, 2002).   Despite the abnormalities and reduced number of these pericytes, tumour blood vessels are seemingly able to tolerate a substantial reduction in normal pericyte coverage.  However, this may imply that even a small number of pericytes can play a large role in promoting vessel stability and tumour survival. For example, tumours with a high degree of pericyte negative vessels appear more vulnerable to VEGF withdrawal due to selective elimination of these vessels  (Benjamin, Golijanin, Itin, Pode, & Keshet, 1999).  In experiments with PDGF-β, whose overexpression has been shown to increase pericyte coverage in tumours (Guo, et al., 2003), blocking PDGFR-β signaling resulted in detachment of pericytes, regression of blood vessel growth, and stabilization of tumour growth (Bergers & Song, 2005).  When targeting both pericytes and endothelial cells together through inhibition of VEGF and PDGF-β, tumour size and vasculature were diminished in a synergistic manner  (Hasumi, Kłosowska- Wardega, Furuhashi, Ostman, Heldin, & Hellberg, 2007).  Such studies highlight the importance of pericytes as vascular targets alongside endothelial cells.   12 1.2.4 Tight junctions  The epithelium forms a cellular barrier that lines tissue compartments and mediates exchange of molecules with the external environment, including gastrointestinal and circulatory uptake.  This barrier consists of junctional complexes between adjacent epithelial cells that confer cell-to-cell adhesion, supply channels for ion exchange, and help regulate paracellular permeability (Matter & Balda, 2003).  Junctional complexes include gap junctions, desmosomes, adherens junctions, and tight junctions (Figure 1.3).   Tight junctions (TJ) are the most apical junctional complex of the epithelial layer.  They are composed of a branching network of integral membrane protein strands that connect adjacent cells, bringing neighbouring plasma membranes into very close proximity and eliminating the intercellular space  (Tsukita, Yamazaki, Katsuno, & Tamura, 2008).  Using transmission electron microscopy (TEM) and freeze fracture experiments, TJ were observed to form a continuous structure that completely circumvents each epithelial cell  (Farquhar & Palade, 1963).  These structural aspects of TJ provide epithelial layers with two important features known as “gate” and “fence” functions.   TJ function as a “gate” by acting as a semi-permeable barrier that restricts the diffusion of molecules based on their charge and size  (Cereijido, Robbins, Dolan, Rotunno, & Sabatini, 1978), leaving passage through the cell membranes via diffusion or active transport as the primary means of epithelial uptake.  Secondly, TJ act as a “fence” because the network of TJ strands prevents mixing of the lipids and proteins of the  13 external plasma membrane from passing between the apical and basolateral surfaces  (Dragsten, Blumenthal, & Handler, 1981).  This is vital for preserving the polarity of the cell and allowing specialized transport at each surface to be maintained.  TJ are also present in endothelium where they help regulate molecular and inflammatory cell permeability, but the majority of endothelial blood vessels do not form the tight barrier seen in epithelial layers.  Only in highly specialized vasculature, like those of the blood-brain-barrier and the blood-retina- barrier, are TJ well developed and able to strictly regulate permeability  (Dejana, Lampugnani, Martinez-Estrada, & Bazzoni, 2000).  Molecularly, the structure of TJ can be separated into three distinct regions: (i) integral membrane proteins that provide mechanical adhesion to adjacent cells, including the occludins, claudins, and junctional adhesion molecules (JAM); (ii) the internal plaque anchoring proteins that link the transmembrane molecules to the cell cytoskeleton, including the zonula occludins (ZO) and membrane associated guanylate kinase (MAGUK) protein families; and (iii) TJ-associated proteins such as α-catenin and cingulin.   Tumour metastasis is believed to occur in a stepwise process  (Nicolson, 1988) in which the presence of TJ potentially impacts several stages.  In cancers of epithelial origin, which comprise approximately 90% of all cancers, the loss of TJ cell-to-cell adhesion of the tumour-containing epithelium is widely accepted to be necessary for invasion of the stroma surrounding the tumour  (Mori, Sawada, Kokai, & Satoh, 1999).  Intravasation of cancerous cells into nearby vasculature  14 would also require bypassing the barrier function of TJ in the vascular endothelium  (Martin, Mansel, & Jiang, 2002).  Finally, extravasation from the circulation into the surround normal tissue, thereby circumventing TJ of local vasculature, is required to establish a metastatic secondary site.  Of these, the initial loss or reduction of TJ complexes in the tumour epithelium is considered the most critical and has attracted the greatest focus of study  (Martin & Jiang, Biochim Biophys Acta, 2008).  One such experiment revealed that enhancing TJ function reduced penetration of tumor cells through mesothelial cells  (Tobioka, Sawada, Zhong, & Mori, 1996).  Conversely, the reduction of TJ in epithelial tissue correlated with tumour progression in rat mammary carcinoma  (Ren, Hamada, Takeichi, Fujikawa, & Kobayashi, 1990).  In a comparison between normal rat colon epithelia and colon tumours, tumour epithelia possessed lower barrier function as shown by decreased transepithelial resistance (TER) and paracellular influx rates  (Soler, Miller, Laughlin, Carp, Klurfeld, & Mullin, 1999). Treatment with hepatocyte growth factor (HGF), a cytokine secreted by stromal cells and associated with tumour progression, reduced TJ expression and function in breast cancer cell lines at both the mRNA and protein levels  (Martin T. , Watkins, Mansel, & Jiang, 2004), adding evidence to the potential importance of TJ in the prevention of metastasis in breast cancer.  1.3 Hypothesis and objectives  The goal of this research was to explore the role of the abnormal solid tumour microenvironment in limiting the penetration and/or distribution properties  15 of the monoclonal antibody therapy trastuzumab.  We hypothesized that the abnormal vasculature of the tumour microenvironment, coupled with the inherently large size of the IgG molecule, would result in populations of cells receiving reduced or incomplete drug exposure, and secondly, that the irregular components of the tumour microenvironment can be modified to alter trastuzumab’s distribution in sold tumours.  The objectives of my research project were: (i) To characterize the extent of trastuzumab penetration and distribution in the MCF7HER2 mammary carcinoma mouse xenograft model; (ii) To utilize the anti-angiogenic effects of imatinib in combination with trastuzumab to improve trastuzumab’s distribution in the xenograft model; and (iii) To identify the localization of tight junctions in MCF7HER2 multilayered cell cultures and observe their effect on trastuzumab penetration in vitro.  The first objective was completed by using immunohistochemistry to visualize the HER2 receptor, bound trastuzumab, CD31 positive blood vessels, and perfusion by either Hoechst33342 or DiOC7 staining.  Comparisons were made between trastuzumab and fluorescein isothiocyanate (FITC)-dextrans of varying molecular weights.  The second aim was achieved by altering pericyte expression in mouse xenografts through administration of imatinib, a receptor  16 tyrosine kinase inhibitor that blocks PDGFR-β signaling.  Changes in pericyte coverage, visualized via the smooth muscle cell intermediate filament desmin, modifications to CD31 positive vasculature, and subsequent trastuzumab penetration effects were analyzed in the MCF7HER2 xenograft model by IHC.  The last objective was accomplished through visualization of junctional complexes by TEM of the luminal surface of multilayered cell cultures (MCC).  Further visualization of TJ by IHC staining of the plaque anchoring protein zonula occludens 1 (ZO-1) was used to confirm the TEM findings and assess localization of trastuzumab treatment in the MCC.   17 Figure 1.1 Ductal carcinoma formation in breast Breast cancers beginning in either the breast lobules or ducts progresses to a ductal carcinoma in situ.  Modification of the surrounding extracellular matrix, loss of junctional complexes, and oncogene activation can all contribute to invasion of surrounding tissue outside of the original lobule or duct.  Cancer cells accessing blood vasculature and setting up secondary sites around the body results in metastatic disease.      Figure reproduced from Love, S.M. (2005)  18 Figure 1.2 Pericytes and endothelial cells A blood vessel cross-section depicting the close association that pericytes have with endothelial cells in the microvasculature.  Pericytes play important roles in vessel stability, regulation of blood flow, angiogenesis, and vascular development.  In tumours, they are more loosely attached to endothelial cells but they are still believed to be important targets for vascular therapy in the tumour microenvironment.       Figure reproduced from Bergers, G., & Song, S. (2005)  19 Figure 1.3 Junctional complexes Tight junctions are the most apical junctional complex and they serve two important functions: (i) to act as a barrier along the paracellular pathway, and (ii) to segregate the lipids and integral membrane proteins of the apical and basolateral portions of the epithelium, thus maintaining polarity of the cells. Adherens junctions lie below tight junctions and serve mainly a structural role in connecting adjacent cells.  Desmosomes and hemidesmosomes are spot-like adhesions that connect cells laterally or to the basement membrane, respectively.  Gap junctions directly connect the cytoplasm of two cells, allowing molecules and ions to be exchanged.     Figure reproduced from Mori, M., Sawada, N., Kokai, Y., & Satoh, M. (1999)  20 2 MATERIALS AND METHODS   2.1 Monolayer cell culture MCF7 human breast adenocarcinoma cells possessing a stably transfected 4.7 kb pRK5 expression plasmid with (MCF7HER2) or without (MCF7NEO) a full length HER2 cDNA were provided by Dr. Wieslawa Dragowska (British Columbia Cancer Agency, Vancouver, BC) and originated from Dr. Alaoui-Jamali (McGill University, Montreal, QB).  HER2-overexpressing cell lines were maintained with 100 µg/mL Geneticin (Invitrogen, Burlington, ON).  Prior to any cell harvest for tumour implantation or multilayered cell culture (MCC) seeding, cells were cultured without Geneticin for 1 week.  MCF7 cell lines were maintained in RPMI1640 (RPMI; Invitrogen) supplemented with 10% FBS (HyClone, Logan, UT), 2 mM L-glutamine (Sigma Chemical, Oakville, ON) and passaged every 3 to 4 days.  Cells were maintained at 37°C and 5% CO2 in a humidified atmosphere.  2.2 Verification of surface HER2 expression by flow cytometry The EPICS ELITE ESP flow cytometer (Beckman-Coulter, Miami, FL) with an Innova Enterprise 621 laser (Coherent, Santa Clara, FL) was used for analysis.  MCF7HER2 and MCF7NEO cells were harvested with trypsin-EDTA and washed once with RPMI media.  Samples of 5x105 cells were blocked using phosphate buffered saline (PBS) supplemented with 1% bovine serum albumin (BSA) and 20% human serum (British Columbia Cancer Agency) for 10 minutes  21 on ice.  Samples were then incubated on ice for 45 minutes with Neu 24.7-FITC antibody (BD Biosciences, San Jose, CA) to detect surface HER2 expression or mouse IgG1-FITC (Caltag Laboratories, Burlingame, CA) as an isotype control. Samples were washed twice with cold PBS plus 1% BSA (PBSB) and left on ice prior to flow analysis.  Quantification was performed using Quantum 24 or 25 FITC premixed beads (Flow Cytometry Standards, San Juan, Puerto Rico) to calibrate the flow cytometer scale for molecules of soluble fluorochrome (MESF). A median relative fluorescence intensity (MRFI) value was calculated for each bead population using Expo-MFA software (Beckman-Coulter).  A calibration curve was constructed and MRFI data from the analyzed sample was converted to MESF values.  Data was corrected for nonspecific fluorescence of the isotype control.   2.3 Tumour xenograft model NOD/SCID mice were bred and housed locally in the Animal Research Centre (ARC) using an individually ventilated caging system.  Female NOD/SCID mice, aged 10-12 weeks, were implanted with 17-ß-estradiol tablets (60-day release, 1.7 mg/tablet; Innovative Research of America, Sarasota, FL) one day prior to inoculation with cells to enhance the growth of the estrogen receptor positive MCF7 cell lines.  Estradiol tablets were inserted through an incision in the dorsal midline of the neck and, using forceps, placed approximately 2 cm beyond the incision site on the lateral side of the mouse.  Incisions were closed with small wound clippers and removed after 10 days.  For each tumour, 1x107  22 MCF7HER2 or MCF7NEO cells in 0.1 mL RPMI media was administered subcutaneously to the sacral region of the mouse.  Tumours were measured using calipers and calculated by three orthogonal diameters (a,b,c) using the formula V=π/6(abc).  Once tumours reached 100-150 mm3 in volume, mice were treated with either 4 mg/kg trastuzumab or sterile saline by intraperitoneal (IP) injection.  Tumours were excised 3-120 hrs after trastuzumab (Hoffman-La Roche, Mississauga, ON; collected from British Columbia Cancer Agency Pharmacy) administration.  For multiple treated groups, additional trastuzumab was delivered at 48 hrs and 96 hrs time points following the initial dose and tumours were excised after 120 hrs. Two hours prior to tumour excision, all mice were dosed IP with 1500 mg/kg 5-bromo-2-deoxyuridine (BrdUrd; Sigma Chemical), a thymine analog that is incorporated into replicating DNA and is used as a marker of proliferation.  Five or twenty minutes prior to excision, 75µl of 0.6 mg/mL of the fluorescent dye DiOC7 (Molecular Probes, Eugene, OR) dissolved in 75% (v/v) DMSO:25% sterile H2O or 50 µl of 20 mg/mL Hoechst33342 (Sigma- Aldrich, St. Louis, MO), respectively, was delivered intravenously (IV) as a marker of blood vessel perfusion.  In experiments testing different molecular weight dextrans, 100 µl of 60 mg/mL 20 kDa, 70 kDa, 250 kDa, or 2000 kDa FITC-Dextrans (FD20S, FD70S, FD250S, and FD2000S; Sigma Chemical) was administered IV twenty minutes prior to excision.  For imatinib treated groups, commercial 400 mg imatinib mesylate tablets (Novartis, East Hanover, NJ) were crushed into powder and dissolved in sterile saline on the day of treatment.  Mice were dosed orally for 7 days, with 100 mg/kg imatinib or saline, using stainless  23 steel 20 gauge curved gavage needles; tumours were excised on day 8 of treatment. Following excision, tumours were immediately frozen at -20°C on an aluminium block, covered in embedding medium (OCT; Tissue-TEK, Torrance, CA) and stored at -20°C until sectioning.  All animal protocols were approved by the University of British Columbia Animal Care Committee and all studies were performed according to Canadian Council of Animal Care guidelines.  2.4 Immunohistochemisty MCC and tumour xenografts were cut into 10 µm cryosections using a Cryostar HM560 (Microm International GmbH, Walldorf, Germany) and air-dried over night.  Sections were fixed in 50% (v/v) acetone/methanol for 10 minutes at room temperature.  Systemically administered, bound trastuzumab was visualized using an Alexa546 anti-human secondary antibody (Invitrogen). Following trastuzumab imaging, the HER2 receptor was detected by staining remaining unbound HER2 with 2.2 µg/mL trastuzumab.  A second cycle of Alexa546 anti-human secondary revealed both HER2 initially bound by systemically trastuzumab treatment and HER2 bound by in vitro trastuzumab staining.  Endothelial cells were identified using an antibody against PECAM/CD31 (rat-anti-CD31; BD Biosciences) and a fluorescent Alexa647 anti- rat secondary antibody (Invitrogen).  Pericytes were detected using a monoclonal antibody against the smooth muscle cell marker desmin (clone D33; Signet Antibodies, Dedham, MA) and a fluorescent Alexa488 anti-mouse secondary antibody (Invitrogen).  Slides stained for tight junction markers were treated with  24 an antibody against ZO-1 (mouse anti-human ZO-1, clone 1/ZO-1; BD Biosciences) and visualized using an Alexa488 anti-mouse secondary (Invitrogen).  To detect proliferating cell BrdUrd incorporation, slides were washed with distilled water for 5 minutes, incubated with 2 M HCL at room temperature for 1 hr, and neutralized with 0.1M sodium borate for 5 minutes. Slides were then treated with a primary antibody against BrdUrd (rat anti-BrdUrd, clone BU1/75; Sigma Chemical) followed by an anti-rat horse radish peroxidase (HRP) conjugate (Sigma-Aldrich) and a metal-enhanced 3,3’-diaminobenzidine substrate (Pierce, Rockford, IL) for 10 minutes.   Lastly, slides were counterstained with hematoxylin, dehydrated, and mounted using Permount (Fisher Scientific, Fair Lawn, NJ) prior to bright field imaging.  2.5 Image acquisition  Images were acquired using a robotic fluorescence microscope (Axio Imager Z1; Zeiss, Toronto, ON) and a cooled, monochrome CCD camera (Retiga 4000R; QImaging, Surrey, BC) utilizing a motorized slide loader and x-y stage (Ludl Electronic Products, Hawthorne, NY).  Customized NIH-ImageJ software (http://rsb.info.nih.gov/ij/), a private domain program developed at at the NIH, was run on a G5 Macintosh computer (Apple Computers, Cupertino, CA).  This system allowed for automated tiling of adjacent microscopic fields, enabling images of entire tumour sections of up to 1 cm2 to be captured.  Images were taken at a resolution of 0.75 µm per pixel.  Composite images of the molecular  25 markers for each tumour were overlaid and aligned using a customized NIH- ImageJ hyperstack function.  2.6 Image analysis Preparation and analysis of composite tumour images were completed using the NIH-ImageJ software and user-supplied algorithms.  To prepare for image analysis, hematoxylin bright field images were used to manually outline tumour borders, remove necrotic tissue sections, and designate the orientation of MCC tissue disks.  Artifacts resulting from staining and imaging procedures were also removed from all layers.  To obtain percentage of positive staining data, positive pixels exceeding a threshold intensity above background fluorescence was calculated as a proportion of whole tumour tissue.  Perfusion data was obtained by creating CD31-positive objects representing individual blood vessels and assessing double-positive staining with the DiOC7 perfusion marker.  A minimum of 20% overlap of blood vessels was required to indicate perfusion. Similar double-positive staining was used to designate vessels positive for pericyte coverage.  To obtain the distribution of marker staining relative to the nearest blood vessel, the average intensity of positive pixels was collected in 1.5 µm increments radiating from central CD31-positive vasculature.  Distance data was collected and averaged over all vessels of each tumour.  In MCC sections, distance analysis was performed similarly, but staining intensity was collected relative to the basolateral side of the layered culture.  To calculate the fraction of necrosis, the number of removed necrotic pixels was compared to the number of  26 pixels constituting the whole tumour image.  Data for all analyses were averaged over a minimum of four tumours.  2.7 Multilayered cell culture Tissue culture inserts (CM 12 mm, pore size 0.4 µm; Millipore, Nepean, ON) were coated with 300 µL of 0.75 mg/mL collagen (rat tail type I; Sigma Chemical, Oakville, ON) and allowed to dry overnight.  Cell monolayers of MCF7HER2 and MCF7NEO were harvested with trypsin-EDTA and 7.5x105 cells in 0.5 mL media were seeded into each insert.  After waiting approximately 15 hrs for the cells to adhere, inserts were transferred to custom built chambers (Figure 2.1) and stirred in 125 mL RPMI media with a 2.5 cm magnetic stir bar at 300 rotations per minute.  Media was supplemented with 1% penicillin/streptomycin (P/S; Invitrogen), 10% FBS, and 2 mM L-glutamine. MCF7HER2 and MCF7NEO cultures were incubated for 5 and 3 days, respectively, to allow for each MCC to reach 150 µm in thickness. Chambers were maintained at 37°C and 5% CO2 in a humidified atmosphere.  Both MCF7HER2 and MCF7NEO cultures were given fresh media after 3 days and 5 days.  2.8 MCC drug treatment and tissue collection Following the exchange of media on the last day of untreated incubation, vessels were dosed with 60 µg/mL trastuzumab or sterile saline.  MCC were incubated in the growth chambers for 1, 6, 24, and 48 hrs to allow for drug penetration into the tissue disks.  Cultures were collected and excess media was  27 removed by dabbing the inside edges of the inserts with KimTech tissue wipes (Kimberly-Clark, Mississauga, ON).  OCT medium was then added and cultures were frozen at –20°C.  To remove the tissue disks from the plastic inserts for sectioning, a custom built coring device was used to cut out the central 8 mm of tissue from frozen samples.  Ejected tissue disks were immediately transported to the cryotome for sectioning.  2.9 Transmission electron microscopy MCC were removed from growth chambers and rinsed twice in minimum essential medium (MEM) before placing in MEM with 2.5% Gluteraldehyde (Ted Pella, Redding, CA) for 10 min at 37°C.  Cultures were then extracted from their inserts before being returned to the 2.5% Gluteraldehyde solution for 1 hr at 37°C.  During this fixation, plastic support membranes detached from the discoid cultures without agitation.  The tissue was then washed three times, 15 minutes each, in sodium cacodylate buffer (Canemco, St. Laurent, QC).  Dehydration was performed in a graded series of acetone, followed by infiltration in Epon (Ted Pella, Redding, CA).  Samples were then embedded and sectioned to a thickness of 55 nm using a Leica UC6 Ultracut microtome (Leica Microsystems Inc, Richmond Hill, ON).  Sections were placed on coated copper grids and stained with uranyl acetate and Sato’s Lead to enhance the contrast of tissue components.  Preparations were viewed using a Tecnai 12 electron microscope (FEI Company, Hillsboro, OR).   28 2.10 Statistical analysis  Data were presented as the mean ± standard deviation (SD).  To assess statistical significance of differences, an unpaired 2-tailed t-test was performed. P values <0.05 were considered significant, as indicated by asterisks.  29 Figure 2.1 Multilayered cell culture growth chamber Custom-built MCC growth chambers were constructed using 150 mL glass jars. Culture inserts with adhered cells were secured in an 8-well plastic cassette and transferred to the growth chamber.  Each well was accessible from both the top and the bottom by media.  125 mL RPMI media supplemented with 1% P/S, 10% FBS, and 2 mM L-glutamine was added to each jar prior to sealing closed and placing at 37°C.  5% CO2 was maintained with tubing fed through the lid.  All components were autoclaved prior to assembly.     Figure adapted from Kyle, Huxham, Chiam, Sim, & Minchinton (2004)  30 3 INVESTIGATING THE ROLE OF PERICYTES IN TRASTUZUMAB DISTRIBUTION  3.1 Rationale and hypothesis  The development of cancer cells into a solid tumour mass is highly dysregulated with imbalances in growth and angiogenesis pathways  (Hanahan & Weinberg, 2000).  The resulting architecture of tumours is vastly unlike normal tissue where the extensive organization of blood vessels is able to consistently deliver components of the circulation to all cells of the body.  Instead, the tumour microenvironment is highlighted by large intervessel distances  (Thomlinson & Gray, 1955), limited and compressed lymphatics  (Helmlinger, Netti, Lichtenbeld, Melder, & Jain, 1997), and a poorly organized and aberrant vasculature (Carmeliet & Jain, 2000).  These abnormalities pose considerable barriers to the delivery of anticancer drugs which, to function ideally, must reach all cells of a cancer capable of regenerating it while at a high enough concentration for cell kill (Minchinton & Tannock, 2006).  Additionally, this disorganization can lead to elevated interstitial fluid pressure  (Flessner, Choi, Credit, Deverkadra, & Henderson, 2005) and intermittent blood flow  (Toffoli & Michiels, 2008) that can also hamper effective drug delivery.  The treatment of solid tumours with systemically administered antibodies is a rapidly growing field in targeted anticancer therapy.  Considerations for the effective penetration of these macromolecules are especially important due to  31 their relatively high molecular weight, antigen binding consumption, rates of internalization, and patterns of tumour antigen expression, all of which can further limit the distribution of antibodies  (Thurber, Schmidt, & Wittrup, 2008). Trastuzumab is one such monoclonal antibody therapy that targets HER2/neu, a growth factor receptor overexpressed in 20% of breast cancers  (Wolff, et al., 2007).  Recent work studying trastuzumab penetration in a mouse xenograft model used a novel method to directly image bound trastuzumab in HER2 overexpressing tumours  (Baker, Lindquist, Huxham, Kyle, Sy, & Minchinton, 2008).  They observed that trastuzumab failed to fully saturate HER2 positive tumour tissue and that there was considerable blood vessel heterogeneity with respect to trastuzumab penetration, leading them to suggest incomplete antibody distribution to be a potential resistance mechanism in solid tumours.  The vasculature of the tumour microenvironment is composed of highly erratic branching patterns with blood vessels that are characteristically dilated and leaky  (McDonald & Baluk, 2002).  In addition to endothelial cell deviations, perivascular cells, or pericytes, are reduced in many tumour types and have been observed to form loose associations with blood vessels  (Morikawa, Baluk, Kaidoh, Haskell, Jain, & McDonald, 2002) (Winkler, et al., 2004).  One treatment strategy that targets these abnormalities involves the concept of vascular normalization, where during anti-angiogenic therapy pro-angiogenic factors are decreased such that immature and nascent vessels are removed, transforming remaining vasculature to a “normal” tissue state and resulting in overall increased  32 blood flow  (Jain, 2005).  Under these conditions, tumours can temporarily experience increased drug delivery and oxygenation that will potentially improve the efficacy of both chemotherapy and radiotherapy.  The role of pericytes during anti-angiogenic therapy is still unclear, but several studies have shown the fraction of pericyte coverage to be increased following VEGFR2 blockade (Winkler, et al., 2004) and anti-VEGF antibody treatment  (Willett, et al., 2004), suggesting selection for the mature fraction of endothelial cells with pericyte association  (Benjamin, Golijanin, Itin, Pode, & Keshet, 1999).  Imatinib is a small molecule inhibitor that prevents phosphorylation by receptor tyrosine kinases, including bcr-abl for chronic myeloid leukemia, and c- kit in gastrointestinal stromal tumours.  Imatinib also interferes with kinase activity of the platelet-derived growth factor receptor beta (PDGFR-β), whose activation is involved in proliferation and survival signaling pathways  (Tallquist & Kazlauskas, 2004).  In MCF7 tumour cells, exposure to imatinib downregulated PDGFR-β and resulted in increased apoptosis and decreased cell viability, growth, and migration  (Rocha, Azevedo, & Soares, 2008).  When imatinib was administered to mice bearing non-small cell lung cancer xenografts, PDGFR-β and VEGF expression was reduced and a decrease in interstitial fluid pressure and increased oxygenation was observed  (Vlahovic, Rabbani, Herndon, Dewhirst, & Vujaskovic, 2006).  Additional analysis revealed reduced microvessel density with a decreased fraction of immature vessels following imatinib treatment, suggesting a normalization effect to be occurring in these tumours  33 (Vlahovic, et al., 2007).  Indeed, pretreatment with imatinib significantly improved the antitumour efficacy of doxorubicin as well as increased tumour concentratration of liposomal doxorubicin.  For the current study, we applied imatinib pretreatment in the MCF7HER2 xenograft model in order to monitor changes in tumour blood vessel properties and to observe their effects on trastuzumab penetration and distribution.  3.2 Results 3.2.1 Extravascular distribution of trastuzumab over time Localization of the HER2 receptor and bound trastuzumab in tumour xenografts was visualized by IHC in mice dosed with 4 mg/kg trastuzumab (Figure 3.1).  Whole tumour composite images are shown at 3 hrs and 8 hrs following trastuzumab dosing in figures 3.1a and 3.1b, respectively.  At early time points (3 hrs), trastuzumab is seen to be extravasating from CD31 positive blood vessels with penetration into the solid tissue within a few cell layers (Figure 3.1c). By 8-24 hrs, percentage of positive trastuzumab staining as a proportion of total tumour tissue shows binding of trastuzumab to the HER2 receptor reaching peak levels (Figure 3.1e). However, despite the extent of distribution at these time points, HER2 positive tissue does not appear completely bound even in regions with high trastuzumab saturation (Figure 3.1d).  One contributing factor to this heterogeneity of distribution appears to be large intervessel distances in these tumours that were seen to reach 200 µm or greater.  Distance analysis of binding around these vessels revealed penetration of trastuzumab to be limited to only  34 125 µm at the 8 hr time points when distribution is highest (Figure 3.1f).  We also observed populations of vessels with much reduced penetration compared to those vessels with ‘maximum’ trastuzumab extravasation.  Clustering of these vessels into distinct regions was observed but the location, area, and extent of reduction was not consistent across the treated tumours.  By 72 hrs, bound trastuzumab in the tumour was greatly reduced at all distances from blood vessels indicating some processing or shedding of the trastuzumab/HER2 complex.  3.2.2 Trastuzumab downregulates HER2 surface expression The distribution of the receptor HER2 over a time course after treatment with trastuzumab was similarly monitored (Figure 3.2).  In saline treated groups, all of the HER2 overexpressing MCF7HER2 cells were still positive for the membrane receptor (Figure 3.2a) despite binding of trastuzumab around the first few surrounding cell layers (Figure 3.1a).  Even by 24 hrs with prolonged and extensive trastuzumab exposure there was no significant change in the proportion of HER2 positive tissue (Figure 3.2c).  However, by 72 hrs HER2 expression was visibly reduced throughout the tumour (Figure 3.2b).  Distance analysis revealed this decrease to be skewed to the receptors located more proximally to the associated blood vessel, cells which would have had the longest and highest amount of antibody exposure (Figure 3.2d).  Disappearance of HER2 during the time course coincides with reduction of bound trastuzumab suggesting simultaneous processing of the coupled components.  35  3.2.3 Trastuzumab induces a growth delay in MCF7HER2 tumours Tumour growth of MCF7HER2 xenografts treated with a single dose of 4mg/kg trastuzumab was compared to untreated or multiple-dosed (0, 2, and 4 day) tumours.  Figure 3.3 shows the inhibition on tumour volume growth produced by trastuzumab treatment.  For the single-dosed group, tumour growth was stalled for the first 5 days after treatment, with average tumour volumes equivalent to pre-treatment size (both 124 mm3, SD +/- 19.4 mm3).  In the multiple-dosed group, tumours experienced a decrease in overall size from 111 mm3 to 85 mm3 over 6 days, a drop of 23% (SD +/- 9.7 mm3).   After 6 days without trastuzumab treatment, untreated control tumours increased in sized by 62% from 124 mm3 to 201 mm3 (SD +/- 9.5 mm3).  3.2.4 Distribution of different MW dextrans relative to vasculature is size- dependent  To assess the effects that molecular size has on drug distribution, a range of FITC-dextrans were administered to tumour bearing mice.  Dextrans of varying molecular weights in MCF7HER2 tumours are shown in Figure 3.4.  The 20 kDa dextran staining was diffuse with low intensity (Figure 3.4a).  Larger dextran sizes of 70 kDa and 250 kDa showed increasingly localized staining in or around blood vessels with higher intensity (Figure 3.4b and 3.4c).  This trend continued for the 2000 kDa dextran where FITC was detected in mostly tight bands around CD31 positive vessels (Figure 3.4d).  The graph showing distribution of dextrans  36 relative to the nearest blood vessel was consistent with these observations of vessel localization (Figure 3.4e).  The 20 kDa dextran showed little accumulation within or near blood vessels whereas the 2000 kDa dextran had the highest intensity of staining close to vessels but also had the sharpest drop in intensity as measurements were taken deeper into the tissue.  Total percent positive tissue for each size of dextran is shown in figure 3.4f.  The high fraction of 20 kDa staining corresponded with the diffuse pattern observed in the sections. Remaining high molecular weight dextrans all displayed reduced fractions of staining compared to the 20 kDa dextran but significant differences between them were not manifest.  3.2.5 Binding of HER2 by trastuzumab increases dextran delivery to tumours  In an unanticipated effect of trastuzumab treatment, we observed that MCF7HER2 tumours that were pretreated with trastuzumab for 4-6 hrs displayed large increases in 20 kDa, 70 kDa, and 250 kDa dextran perfusion compared to untreated or 2 hr treated tumours (Figure 3.5).  Dextrans of all three sizes measured displayed increased average intensity of staining that was greatest at 4 hrs of trastuzumab treatment but decreased by 6 hrs (Figure 3.5a). Additionally, the 70 kDa dextran appeared to undergo the greatest increase in staining.  Tumour images with FITC represented in grey highlight the difference in intensity between 2 hr and 4 hr treated tumours for the 70 kDa dextran (Figure 3.5b and 3.5c).  Control groups consisting of MCF7NEO tumours treated with  37 trastuzumab for 4 hr or MCF7HER2 tumours treated with a FITC-IgG isotype control for 4 hr were compared to untreated and 4 hr trastuzumab treated MCF7HER2 groups (Figure 3.6a). 24 hr and 120 hr trastuzumab pretreatment were also tested for 70kDa dextran delivery (Figure 3.6b).  All controls failed to exhibit similar increases in dextran staining compared to 4 hr trastuzumab pretreatment of MCF7HER2.  Distance distribution analysis of 70 kDa dextran relative to vessels showed that the increase in intensity occurred at all distances from CD31- positive vasculature (Figure 3.7a).  Interestingly, the fraction of vessels positive for dextran, as determined by double-positive analysis of FITC and CD31, was seen to increase at 4 hrs and 6 hrs trastuzumab treatment, indicating a possible increase in vascular permeability to the 70 kDa dextran (Figure 3.7b).  3.2.6 Increased necrosis, decreased MVD, and decreased pericyte coverage following imatinib treatment Imatinib, a PDGFR-β inhibitor shown to reduce pericyte recruitment to blood vessels, was administered to tumour bearing mice.  Tumour and blood vessel analysis of mouse xenografts with and without 7-days imatinib treatment is shown in Figure 3.8.  Necrosis, which is normally found at very low levels in the MCF7HER2 tumours, was significantly increased by treatment with imatinib (Figure 3.8a).  In remaining viable tissue area, the microvessel density was decreased as shown by percentage staining of CD31 positive pixels (Figure 3.8b).  Additionally, the number of vessels per unit area was also decreased in the treated group (Figure 3.8f).  These results are consistent with similar  38 observations in NSCLC xenograft tumours.  Average vessel size also appeared to be decreased although not significantly so (Figure 3.8c), indicating that smaller vessel size may also be contributing to the lower CD31 staining.  Analysis of pericytes by desmin staining revealed a decrease in desmin positive tissue after treatment with imatinib (Figure 3.8d).  Since the presence of pericytes is typically reliant on the number of blood vessels, a portion of the desmin decrease may be attributed to the drops in vessel number and possibly vessel size.  However, we found that the remaining vessels in the imatinib treated group displayed a lower proportion of vessels positive for desmin when compared to untreated tumour vessels (Figure 3.8e).  3.2.7 Trastuzumab is decreased in imatinib treated tumours Groups with and without imatinib treatment were administered trastuzumab for 4hrs to compare microenvironmental effects of imatinib on trastuzumab’s extravascular penetration and whole tumour distribution (Figure 3.9).  Overall staining of bound trastuzumab decreased in mice pre-treated with the imatinib regime (Figure 3.9b).  There was, however, no significant change in levels of the small molecule perfusion dye, DiOC7, between treated and untreated groups (Figure 3.9a).  Vessel distance analysis revealed that decreases in trastuzumab binding were present at all distances from blood vessels (Figure 3.9c) and not restricted to specific regions such as distal from vasculature.  Whole tumour tissue maps of these tumours were created and highlighted regions are shown in figures 3.9d and 3.9e.  Presence of bound  39 trastuzumab was visibly decreased in imatinib treated tumours.  Interestingly, there was an increased occurrence of vessels with minimal or absent trastuzumab staining (Figure 3.9e, arrows), which could account for the loss of bound antibody at distance both near and far from blood vessels.  3.2.8 Extravascular penetration of trastuzumab is greater in blood vessels positive for pericyte coverage Since the presence of desmin was decreased in tumours treated with imatinib, I performed staining and analysis to identify the localization of blood vessel-associated pericytes with respect to trastuzumab distribution, which was also decreased following imatinib treatment (Figure 3.10).  In composite images of trastuzumab, CD31, and desmin immunohistochemical staining (Figure 3.10a) we found increased trastuzumab distribution around blood vessels that were also positive for pericytes.  For blood vessels negative for the desmin stain, trastuzumab penetration was visibly decreased or absent.  These observations were consistent for tumours in both the imatinib treated and untreated groups. When comparing the distance distribution of trastuzumab between desmin positive and negative vessels (Figure 3.10b), I found that the data separated into two distinct profiles.  Desmin positive vessels did indeed demonstrate greater trastuzumab staining at all distances when compared to total vessel averages. For desmin negative vessels, values were considerably lower than total vessel averages and at distances beyond 50 µm trastuzumab staining was severely reduced.  40  3.3 Discussion  Utilizing the method of directly visualizing systemically administered tumour bound trastuzumab antibody, we have successfully monitored trastuzumab penetration and distribution in a MCF7HER2 breast cancer xenograft model.  Profiles for extravascular staining were consistent with previous findings involving xenografts created from MDA-435/LCC6HER2  (Baker, Lindquist, Huxham, Kyle, Sy, & Minchinton, 2008), a cell line that, although recently identified as having a melanoma origin by genetic profiling  (Rae, Creighton, Meck, Haddad, & Johnson, 2007), was the first study to observe the in vivo extravastion of trastuzumab in a HER2-overexpressing tumour model.  Trastuzumab functions primarily through binding of the HER2 receptor (Nahta, Yu, Hung, Hortobagyi, & Esteva, 2006).  Once bound, trastuzumab works through multiple mechanisms that include either attacking the tumour cells by initiating an antibody-dependent cellular response  (Cooley, Burns, Repka, & Miller, 1999) or by reducing HER2 through internalization/degradation or downregulation, resulting in a blockage of downstream growth signaling  (Yakes, Chinratanalab, Ritter, King, Seelig, & Arteaga, 2002).  In studies performed by Warbuton et al, MCF7HER2 xenografts were treated with trastuzumab and HER2 levels were found to be decreased in flow cytometric analyses  (Warburton, et al., 2004).  Through immunohistochemical staining of HER2, we observed a reduction of HER2 protein 72 hrs after exposure to trastuzumab.  Disappearance  41 of HER2 occurred proximal to blood vessels, in cells that were first exposed to trastuzumab and bound for the longest duration.  Because immunohistochemical staining for both HER2 and trastuzumab did not reveal any cytoplasmic localization (data not shown), this indicates removal of HER2 by either internalization or, possibly, shedding of the coupled complex.  Treatment with trastuzumab also resulted in a growth delay of tumour volume compared to control mice.  Interestingly, the inhibition of growth was manifest prior to the detection of HER2 reduction, suggesting an additional suppressive role of trastuzumab prior to downregulation of HER2.  Also of note, staining of the incorporated proliferation marker BrdUrd remained unchanged throughout the time course.  If changes in proliferation occur downstream from HER2 reduction, we may not be able to detect such phenotypic effects in the tumour microenvironment without longer treatment durations.  As expected, extravascular penetration of trastuzumab originated from perfused blood vessels and the stained tissue fraction increased over time. However, in both the MCF7HER2 and MDA-435/LCC6HER2 cell lines, prolonged exposure of tumours to circulating trastuzumab resulted in a heterogeneous distribution that failed to bind all tumour cells despite constitutive HER2 expression.  Observation of neighbouring blood vessels revealed inconsistent intervessel distances, with cells distant (>150 µm) from the nearest vessel being out of the range of trastuzumab penetration.  This reduction of vascular density has been well documented and is believed to be a distinctive barrier in tumours,  42 characteristic of the highly dysregulated growth of tumour cells and their vasculature.  Compounding this problem, we also observed populations of blood vessels with reduced trastuzumab that was limited to only a few cell layers or altogether absent.  Analysis across entire tumour cross-sections showed these vessels to occur in clustered regions that could span over half the tumour area but were often interspersed between regions of trastuzumab positive vessels. Decreased penetration from such vessels may have resulted from numerous sources, including variations in intersitial fluid pressure, intermittent blood flow, and composition of the extracellular matrix.  Additionally, the stability, maturity, and extent of developing blood vessel architecture can also influence penetration of drugs  (Hormigo, Gutin, & Rafii, 2007).  Perhaps the transiently increased blood flow achieved through vascular normalization strategies can improve the overall consistency of trastuzumab distribution.  However, the proper timing of combining trastuzumab with anti-angiogenic agents would be critical and factors such as the large intervessel distances would still be a concern.   A range of FITC-dextrans was applied to the MCF7HER2 tumour xenograft model to gain an understanding of the effect molecular size has on extravascular penetration and whole tumour distribution.  This is relevant because of the inherent, relatively large size of the trastuzumab molecule in comparison to conventional chemotherapy agents.  As an IgG, trastuzumab has a molecular weight of 150 kDa while many traditional cancer treatments are in the size range of hundreds of Daltons.  The size range chosen for this experiment consisted of  43 the 20 kDa dextran, which can represent small nutrient molecules with high permeability, the 70 kDa and 250 kDa dextrans, which are useful for evaluating endothelial permeability changes, and the 2000 kDa dextran, which is assumed to be purely intravascular.  Interestingly, the molecular radius of trastuzumab is closest to the 70 kDa FITC-dextran (5.23 and 6.4, respectively; 20 kDa and 150 kDa are 3.2 and 8.25, respectively)  (Ambati, et al., 2000) because the 70 kDa dextran experienced the greatest increase in staining following trastuzumab treatment.  This result suggests that the increased delivery of dextran may be due to altered vascular permeability that is accessible to sizes similar to that of an IgG.  Furthermore, control groups showing only increased dextran delivery in 4 hr treated MCF7HER2 compared to MCF7NEO and FITC-IgG groups points to the actual binding of HER2 by trastuzumab as the initializing event.  Lack of increased dextran delivery in longer trastuzumab treatments provides further insight that this effect is an acute event shortly following immune complex formation.  One study, which used the reverse passive Arthus (RPA) model to induce immune complex formation in mouse muscle, measured changes in 70 kDa FITC-dextran permeability and found that rapid changes in microvascular permeability followed  (Lister, James, & Hickey, 2007).  These changes were accompanied by induction of complement-dependent leukocyte recruitment, neutrophil-dependent microvascular dysfunction, and platelet-endothelial interactions consistent with acute inflammatory response initiation.  Our findings of an increased fraction of vessels positive for 70 kDa dextran after 4 hr trastuzumab dosing may represent a similar increase in permeability.  In addition  44 to HER2 downregulation, induction of an inflammatory response would present an additional downstream effect of trastuzumab binding that can be visualized in the MCF7HER2 xenograft model.  Imatinib mesylate has been shown to have vascular effects in NSCLC tumour xenografts through PDGF-β/PDGFR-β receptor tyrosine kinase inhibition (Vlahovic, et al., 2007).  Treatment with imatinib was shown to decreased VEGF and microvessel density, resulting in proportionally decreased CD105 endothelial cell immaturity, as well as decreased interstitial fluid pressure and improved oxygenation  (Vlahovic, Rabbani, Herndon, Dewhirst, & Vujaskovic, 2006).  They proposed that normalization of NSCLC tumour vasculature contributed to their observed increase in delivery of the small molecule chemotherapeutic docetaxel. Imatinib pretreatment of our MCF7HER2 xenografts resulted in similar reductions of microvessel density and tumour growth inhibition.  We also observed an overall decrease in pericyte positive tissue as well as a decline in the fraction of vessels associated with pericytes as shown by desmin staining.  Although these results are not unexpected since PDGF-β signaling is known to play an important part in pericyte recruitment  (Ostman, 2004), the role of imatinib as a normalizing agent thus appears conflicting as it is able to exert an anti-angiogenic effect on immature vessels but it is also active in blocking recruitment of the perivascular cells required for protection and stabilization of remaining endothelial vessels. Nonetheless, the timing for these events is still largely unknown; pericytes, for example, have been observed to increase before undergoing acute reduction  45 when treated over long periods with anti-angiogenic factors  (Inai, et al., 2004).  A precise temporal understanding of the anti-angiogenic effects of a drug such as imatinib will reveal windows of time where combinatorial therapy can be utilized.  Because of the critical role that pericytes play in the normal development of microvessel formation during development, it has been proposed that the heterogeneity of pericyte coverage between and within tumours may correlate with the structural irregularities and functional heterogeneity found in tumour vasculature  (Abramsson, Lindblom, & Betsholtz, 2003).  In one set of experiments, immature vessels lacking pericytes were preferentially destroyed following VEGF withdrawal to tumour xenografts  (Benjamin, Golijanin, Itin, Pode, & Keshet, 1999), suggesting that pericyte positive vessels are less reliant on VEGF for their survival.  For our present study, we observed that when comparing blood vessels with and without close pericyte association, trastuzumab penetration was much greater around those positive for the perivascular cells.  Additionally, regardless of imatinib treatment, blood vessels negative for pericytes displayed consistently reduced or absent trastuzumab penetration.  In the context of the heterogeneous distribution of trastuzumab, especially with regard to those vessels exhibiting sub-optimal penetration, these results offer a partial explanation for this occurrence while using a tangible marker of tumour blood vessels.  However, additional analysis between regions of tumour with high and low trastuzumab saturation will be required to obtain a better understanding of the impact of these perivascular cells.  Of particular  46 interest will be to explore if phenotypic properties of tumour blood vessels, such as pericyte coverage, can simply result in vessels that are preferentially permeable to antibodies and other drug therapies regardless of the surrounding vascular network.  Equally important will be to identify distinct, localized regions within a tumour that contain vasculature with varying phenotypic properties, but equivalent pericyte association, and correlate surrounding trastuzumab distribution.   Further experiments examining pericyte blood vessel localization, specifically in combination with conventional anti-angiogenic agents, will help to assess these roles in influencing the microregional distribution of trastuzumab.  The combination of reduced pericyte levels and a marked decreased in penetration of trastuzumab from vessels negative for pericyte coverage may have contributed to the overall decrease of trastuzumab distribution in MCF7HER2 tumour xenografts following imatinib pretreatment.  Imatinib, which functions in these tumours by interfering with PDGFR-β activation by PDGF-β, plays dual roles in producing anti-angiogenic effects while impairing the recruitment of perivascular cells.  Vascular normalization, by elimination of immature and dysfunctional vessels, is typically believed to increase drug distribution by increasing overall efficiency of remaining blood vessels.  Conversely, studies have shown that normalization decreases the proportion of highly leaky vessels that large molecules, such as immunoglobulins, utilize and therefore decreases their overall permeability  (Tong, Boucher, Kozin, Winkler, Hicklin, & Jain, 2004). As well, the additional effect of imatinib in preventing recruitment of pericytes  47 may have further contributed to this reduction of trastuzumab distribution.  The occurrence of both these actions is somewhat contradictory, however, since pericyte negative vessels are thought to be less stable and are more likely to be represented in the leaky vessel population that would normally be pruned during anti-angiogenic therapy.  It is possible that there are inconsistencies in classifying leakiness of these vessels, especially in relation to tumour pericyte expression where coverage is often incomplete and only loosely associated to blood vessels. Future experiments affecting vascular remodeling in the absence of reduced pericyte recruitment, or experiments increasing the extent of pericye coverage, will provide insight into the optimal vascular microenvironment for trastuzumab- based therapies.   48 Figure 3.1 Immunohistochemical imaging of bound trastuzumab in trastuzumab treated MCF7HER2 tumour xenografts NOD/SCID mice were dosed with 4 mg/kg trastuzumab (brown) for 0-72 hrs. Whole tumour images are shown following 3 hrs (a) and 8 hrs (b) trastuzumab treatment, with insets (c, d).  Staining for CD31-positive endothelial cells (blue), HER2 (green), and perfusion around vessels via DiOC7 (cyan) are also shown. Percentage of trastuzumab positive tissue was low at 3hrs and reached a maximum at 8-24 hrs (e); n = 6 per group.  Distribution of trastuzumab relative to blood vessels (f) revealed increased intensity within 25 µm of vessels and an overall greater intensity of staining at all distances after 8 hrs compared to 3 hrs treatment.  Error bars denote mean ± SD.      49 Figure 3.2 HER2 expression in MCF7HER2 xenografts following treatment with 4 mg/kg trastuzumab HER2 receptor levels (green) were visualized in whole tumour sections that were treated with saline (a) or trastuzumab for 72 hrs (b).  The HER2 receptor is visibly reduced after 72 hrs treatment but there was no significant change in HER2 percent positive tissue from 0-24 hrs (c); n = 5 per group.  HER2 distance distribution data for untreated and 120 hrs trastuzumab treated groups (d) revealed that the greatest amount of HER2 reduction occurred at closer distances to blood vessels, perhaps reflecting the tissue that was exposed to trastuzumab for the longest period.  Error bars denote mean ± SD.       50 Figure 3.3 Growth delay of MCF7HER2 xenografts following single or multiple doses of 4 mg/kg trastuzumab Mice harbouring MCF7HER2 tumours were given either a single dose of 4 mg/kg trastuzumab at day 0, 4 mg/kg trastuzumab on days 0, 2, and 4, or saline. Tumours of saline treated mice experienced steady growth while single dosed groups experienced growth arrest.  The multiple-dosed group showed a slight decline in tumour size by 6 days after initial treatment.  Error bars denote mean ± SD.  n = 6 per group.      51 Figure 3.4 Immunohistochemical imaging of different molecular weight FITC-dextrans in MCF7HER2 xenografts FITC-dextrans of different molecular weights were administered to mice 20 min prior to sacrifice (yellow).  HER2 (green) and CD31 positive vessels (blue) are also shown.  Staining of the 20 kDa dextran (a) was diffuse, while the larger sized dextrans (70 kDa, 250 kDa, and 2000 kDa; b-d, respectively) exhibited increasing localization towards blood vessels.  Dextran distribution around blood vessels was consistent with these findings (e); n = 4 per group.  The fraction of dextran positive pixels (f) highlighted the diffuse staining of the 20 kDa dextran compared to the larger sizes.  Error bars denote mean ± SD.      52 Figure 3.5 Increased dextran delivery following 4 mg/kg trastuzumab treatment (a) The average intensity of all sizes of dextrans was greatly increased for 4 hrs treatment but this effect was reduced after 6 hrs.  n = 4 per group.  Whole tumour images of 70 kDa dextran between 2 hrs (b) and 4 hrs (c) trastuzumab signifies the magnitude of increase of dextran distribution (grey).  The influx of dextran is possibly due to increased permeability caused by an inflammatory response to the trastuzumab-HER2 immune complex formation.  Error bars denote mean ± SD.       53 Figure 3.6 FITC-Dextran controls (a) The saline treated group had similar levels of dextran to the IgG isotype control and to the MCF7NEO tumours that do not express the HER2 receptor. These indicate that trastuzumab and HER2 both need to be present to illicit the improved dextran response.  n = 5 per group. (b) Prolonged exposure to trastuzumab beyond 24 hrs did not show increased distribution of dextran, suggesting that this is an acute effect caused by the initial binding of trastuzumab to HER2.  n = 4 per group.  All controls were performed using the 70 kDa FITC- dextran.  Error bars denote mean ± SD.   *P < 0.05.     54 Figure 3.7 Increased blood vessel permeability following 4 mg/kg trastuzumab treatment (a) The dextran distribution around vessels showed not only an increase in 70 kDa FITC-dextran intensity at regions close to vessels, but regions distant from vessels were also greatly increased after 4 hrs and 6 hrs trastuzumab treatment. (b) This graph represents the fraction of CD31-positive blood vessels that were doubly stained with FITC-dextran.  These results suggest that a higher proportion of blood vessels in the 4 hr or 6 hr groups were positive for dextran compared to the 2 hr treated group (positivity in this case representing either blood vessels that contain the dextrans intravascularly or blood vessels that have extravasation of dextran into the surrounding tissue).  Error bars denote mean ± SD.  n = 12 per group.       55 Figure 3.8 Effects of imatinib mesylate on the vasculature of MCF7HER2 xenografts Treatment with the PDGFR-β inhibitor resulted in a large increase in necrosis (a) and decreased microvessel density (b).  The decrease in vasculature may be explained by a combination of overall decreased vessel size (c) and decreased number of vessels (f).  Pericytes, identified by desmin staining, rely on PDGF- β/PDGFR-β signaling for recruitment to endothelial blood vessels.  Imatinib treatment reduced the fraction of desmin positive tissue (d) and also decreased the fraction of vessels positive for pericyte coverage (e).  Error bars denote mean ± SD.  n = 6 per group.  n.s., not significant; *P < 0.05.     56 Figure 3.9 Imatinib treatment decreases trastuzumab distribution in MCF7HER2 xenografts Although there were no noticeable changes in vessel perfusion (a), the fraction of trastuzumab positive tissue decreased in mice dosed with imatinib (b); n = 6 per group.  The distribution of trastuzumab around vessels was markedly decreased at all distances (c).   Tumour sections from imatinib untreated (d) and treated (e) groups were stained for bound trastuzumab (brown), CD31-positive endothelial cells (blue), DiOC7 (cyan), and HER2 (green).  Notably, in the imatinib treated group, there were many perfused vessels that appeared to be almost completely lacking in trastuzumab penetration (e, arrows).  This population of vessels may contribute to the decreased distance distribution of trastuzumab at distances both near and far from vessels.  Error bars denote mean ± SD.  n.s., not significant; *P < 0.05.            57   Figure 3.9 (continued)     58 Figure 3.10 Increased extravascular penetration of trastuzumab from blood vessels positive for pericyte coverage (a) Tumours were stained for bound trastuzumab (red), CD31-positive endothelial cells (blue), and desmin (pink).  Endothelial blood vessel coverage by pericytes is clearly visible in these sections.  The extravascular penetration of trastuzumab is seen predominately around vessels that are closely associated with the pericytes. (b) The trastuzumab distance distribution curves were generated for vessels positive for pericytes, negative for pericytes, and for all vessels.  Desmin negative vessels had some level of regional trastuzumab but these were considerably lower than for the desmin positive vessels.  Error bars denote mean ± SD.  n = 6 per group.     59  4 INVESTIGATING THE ROLE OF TIGHT JUNCTIONS IN TRASTUZUMAB DISTRIBUTION   4.1 Rationale and hypothesis    The incomplete distribution of drugs in solid tumours will result in cells receiving non-optimal drug exposure, reducing the efficacy of anti-cancer treatments.  To function ideally, drugs must not only reach all cells within a tumour that are capable of regenerating it, such as clonogenic or tumour stem cells, but they must also reach these distances at a sufficient concentration to exert their specific mechanism of action  (Minchinton & Tannock, 2006).  One major obstacle to complete drug distribution in solid tumours is the presence of populations of cells that are distant (>100 µm) from functioning blood vessels, which can result from the imbalance between the uncontrolled growth of cancerous tissue and the aberrant development of a vascular network.  Drugs traveling through these relatively large distances are hampered by numerous factors, including: (1) barriers to convective flow, such as elevated interstitial fluid pressure arising from absent or reduced lymphatics  (Helmlinger, Netti, Lichtenbeld, Melder, & Jain, 1997), (2) physicochemical drug properties such as MW/size, solubility, and charge that limit extravascular diffusion rates, (3) drug consumption via target binding, cellular uptake, and metabolism/sequestration, and (4) composition of the extracellular compartment  (Netti, Berk, Swartz, Grodzinsky, & Jain, 2000), which includes the extracellular matrix and other  60 stromal barriers.  Increased understanding of these processes will allow for the development of strategies to improve distribution of drugs in solid tumours.  One component of the interstitium that has interesting implications in tumour biology is that of intracellular junctions.  Tight junctions form rings of ultra- tight protein bands connecting adjacent cells, functioning primarily to regulate paracellular transport and maintain apical/basolateral polarity through localization of lipids and integral membrane proteins.  Many studies have focused on the disruption of tight junction formation as a means to modify protein absorption across the endothelial blood-brain barrier and other epithelial layers  (Sandoval & Witt, 2008)  (Deli, 2009).  The loss or improper formation of tight junctions has been associated with the progression of tumours to the metastatic or invasive state in various cancers  (Martin & Jiang, 2008).  In a spheroid culture model, the differential expression of tight junctions in treated breast cancer cells was used as a parameter for evaluating anti-cancer drug combinations  (dit Faute, Laurent, Ploton, Poupon, Jardillier, & Bobichon, 2002).  In a separate tumour spheroid study, a cell line with known tight junction formation exhibited antibody localization to either the apical (outer) or basolateral (inner) surfaces of the cell, suggesting tight junctions as a means for preventing antibody penetration (Pervez, Kirkland, Epenetos, Mooi, Evans, & Krausz, 1989).  The use and development of antibodies for targeted cancer therapy is now well established.  However, there are additional considerations when assessing  61 the extent of antibody penetration into solid tumour tissue.  Binding affinity of the antibody for its target, degree of tumour expression of antigen, effect of antibody size on diffusion and systemic clearance, and rates of internalization and degradation of the bound immune complex are some major concerns for mathematic modeling of antibody penetration  (Thurber, Schmidt, & Wittrup, 2008).  Recently, the distribution of trastuzumab was demonstrated in solid tumour xenografts in vivo using a novel immunohistochemical method where trastuzumab bound directly to HER2 was visualized using an anti-human fluorescent secondary  (Baker, Lindquist, Huxham, Kyle, Sy, & Minchinton, 2008).  Analysis of trastuzumab distribution around perfused blood vessels revealed bound antibody up to 150 µM into solid tumours.  However, it was observed that staining was non-uniform around vessels and, despite large doses of trastuzumab, there remained unbound HER2 in tumour sections.  Multilayered cell cultures (MCC) are 3-dimensional tissue disks that are used to evaluate the ability of a given substance to diffuse across solid tissue (Kyle, Huxham, Chiam, Sim, & Minchinton, 2004).  Generally, cultured cells are grown on a plastic support membrane disk to a desired thickness, typically 200- 250 µM, before being placed in an apparatus where the disk of tissue separates two reservoirs of media.  Drug can then be added to one reservoir and its flux measured in the receiving medium.  Additionally, MCC can be sectioned after treatment and important information regarding the distribution of fluorescent or labeled drugs can be collected.  For this current study, I employed a modified  62 application of the MCC model whereby trastuzumab had access to both the apical (or ‘luminal’) and membrane-attached (or ‘basolateral’) layers of a culture over-expressing HER2.  Visualized localization of bound trastuzumab was then used to assess the role of tight junctions in the limitation of antibody distribution in solid tumours.  4.2 Results   4.2.1 Luminal and basolateral sides of MCF7HER2 MCC exhibit dissimilar trastuzumab penetration IHC staining of trastuzumab distribution was observed in MCC exposed to trastuzumab for up to 48 hrs in growth chambers (Figure 4.1).  Overall thickness increased by an average of 120 µM due to growth of the cultures.  The formation of domes of cells along the luminal side of the MCC was also observed.  After 1 hr of exposure, trastuzumab was bound several cell layers into the basolateral side of the thick culture but only sparsely along the luminal layer of cells. Trastuzumab penetration increased from the basolateral side of the MCC until complete coverage of HER2 positive tissue by 48 hrs.  Conversely, along the luminal side of the culture, penetration increased at a slower rate and was incomplete.  By 24 hrs, some luminal regions with penetrating trastuzumab displayed staining comparable with the leading edge of penetration from the basolateral side of the MCC.  The average intensity of trastuzumab distribution relative to distance from the basolateral side of the MCC following 6hrs of  63 exposure (Figure 4.2) highlights the penetration differences between the two sides of the culture.  4.2.2 Luminal tight junction localization in transmission electron micrographs Ultrathin cross-sections of MCF7HER2 MCC without trastuzumab treatment were imaged by transmission electron microscopy to examine differences between the luminal and basolateral cell layers  (Figure 4.3).  Cellular junctions located along the luminal side of the MCC contained plasma membranes in very close contact with adjacent cells (Figure 4.3a).  This degree of membrane proximity and the presence of kiss points (Figure 4.3b, arrowhead) are indicative of the presence of tight junctions.  Cell-to-cell adhesion was further enhanced by the presence of desmosomes and cell membrane interdigitation (Figures 4.3c, 4.3d) between the cells of the luminal layer of the MCC.  Electron micrographs of cell junctions along the basolateral layer of the MCC revealed large extracellular spaces between neighboring cells (Figure 4.3e, 4.3f) when compared to the luminal-side cellular junctions.  Junctional complexes were also less abundant, with only occasional desmosome formation and no indication of tight junctions.  Remnants of the collagen layer that the MCC were grown upon were only seen in trace amounts below this basolateral layer and the establishment of any sizable extracellular matrix was not observed anywhere throughout the culture.  64  4.2.3 Tight junctions restrict trastuzumab penetration in MCC To support the morphologic observations of tight junctions via electron microscopy, immunofluorescent labeling of a specific tight junction marker was performed using the anti-ZO-1 antibody (see Figure 4.4).  Electron microscopy findings were confirmed when ZO-1 was found to form a continuous band of staining along the luminal cell layer of the MCC (Figure 4.4a).  Tissue invaginations (Figure 4.4a, arrowheads), which are likely due to the dome-like organization of MCF7HER2 cells, maintained the ZO-1 barrier.  Along the basolateral side of the MCC (Figure 4.4b), ZO-1 staining was virtually absent between cells.  Throughout the middle portion of the MCC, positive staining was detected around small clusters of cells, possibly arising from invaginations originating from outside of the plane of cross-section.  The limited penetration of trastuzumab into the luminal side of the MCC is shown in Figure 4.5.  By overlaying the ZO-1 marker (Figure 4.5b) over bound trastuzumab staining, the distribution of trastuzumab, shown in red, is revealed to be limited to either cells that are completely beyond the ZO-1 barrier or to the portion of cells that are on the outward facing apical/luminal side of ZO-1 positive cells.   65 4.2.4 Tight junctions form discontinuous and haphazard patterns in MCF7HER2 tumour xenografts Immunofluorescent imaging was also performed on tumour xenograft cryosections from mice treated with 4 mg/kg trastuzumab for 24 hrs (see Figure 4.6).  Figure 4.6a shows the distribution of trastuzumab (red) around CD31 positive blood vessels (blue) to be irregular, with some vessels demonstrating high levels of penetration and others with low intensity staining at a distance of only 1 cell layer from the blood vessel (arrowheads).  Tight junctions appeared randomly distributed throughout the tumour (Figure 4.6b) without forming a continuous structure that appears capable of regulating intracellular permeability. At the cellular level, the majority of ZO-1 staining was comprised of either complete rings that encircled whole cells or segments spanning the portion of cell membrane in contact with one or two adjacent cells.  The heterogeneity of ZO-1 staining may be a result of differences in cell orientation, which may be a factor if cells have maintained aspects of polarization.  In the composite image (Figure 4.6c), tight junctions do not appear to form distinct barriers that would function to limit the penetration of trastuzumab from blood vessels.  Additionally, the distribution of trastuzumab did not seem affected by the individual rings or segments of ZO-1 at regions near or distant from blood vessels in these tumour xenografts.  From these results, there was no evidence suggesting that TJ were responsible for the exclusion of trastuzumab from vessels in this xenograft model.   66 4.4 Discussion  The presence of tight junctions in the MCF7 breast adenocarcinoma line was first identified by TEM in clusters of cells forming around cellulose sponges (Russo, Bradley, McGrath, & Russo, 1977).  In our present study, morphological observations by TEM confirmed the localization of tight junctions in MCF7HER2 multicellular layers after immunohistochemical staining revealed the tight junction peripheral plaque anchoring protein ZO-1 to be localized to one side of the thick cultures when grown on a permeable support membrane.  Similar cell models that are used to examine epithelial tight junction properties include human colon CaCO-2 and canine kidney MDCK cell lines.  Both cultures form confluent monolayers that spontaneously differentiate to form tight junctions between cells and are useful for assessing dietary flux of various molecules across intestinal and renal epithelium  (Rossi, Marciello, Sandri, Bonferoni, Ferrari, & Caramella, 2008)  (Arthur, 2000).  Additionally, changes in tight junction barrier integrity following a wide variety of treatments have been assessed through measurements of trans-epithelial electrical resistance (TER or TEER) across the monolayers  (Hughes, Kurth, McGilligan, McGlynn, & Rowland, 2008)  (McCann, et al., 2007).  CaCO-2 cells have also been grown as multilayers but confocal scanning laser microscopy revealed resulting cultures to be variable with tight junctions not restricted to the most apical cell layer  (Marasanapalle, Li, Polin, & Jasti, 2006)  (Rothen-Rutishauser, Braun, Günthert, & Wunderli-Allenspach, 2000) and culture time required for CaCO-2 cells to form useable multilayers was three to four weeks.  MCF7HER2 cells grown in the MCC chamber model  67 developed multilayers with consistent increases to thickness over a period of five to six days.  More importantly, the formation of a continuous, apical localization of tight junctions by these MCF7HER2 cells provides a novel model for studying transcellular and paracellular transport across a multicellular epithelial layer.  Numerous drug penetration studies have examined the structure and composition of the extracellular matrix and have shown its ability to act as a barrier  (Davies, Berk, Pluen, & Jain, 2002) (Hassid, Eyal, Margalit, Furman- Haran, & Degani, 2008) (Marasanapalle, Li, Polin, & Jasti, 2006) or a channel for conducting extravascular fluids, such as in vasculogenic mimicry (Maniotis, et al., 1999) (Folberg, Hendrix, & Maniotis, 2000) (Clarijs, Otte-Höller, Ruiter, & de Waal, 2002).  For our MCC, rat tail type I collagen (225 µg per insert) used to initially seed the MCF7HER2 cells was found to be almost completely absent in TEMs after 5 days of growth.  Only small remnants of collagen were detected below or between the basolateral layer of cells.  The majority of the initial collagen was likely utilized during adherence to the permeable membrane and/or multilayer development.  Since no other ECM component was detected in the MCC, and with the absence of cellular constituents such as adipocytes and fibroblasts, we were unable to assess stromal impact on penetration in this experimental model.    Cell-packing density is another property of tumors that has been investigated for its effect on drug penetration  (Grantab, Sivananthan, & Tannock, 2006).  Granteb et al compared multicellular layers of different cell lines and showed an inverse correlation between higher packing density and drug  68 penetration.  Additionally, they reported that toxicity to the chemotherapeutic drug doxorubicin was increased in cell lines with loosely packed cells.  These results are consistent with our TEM images showing highly packed cells present throughout the upper portion of our MCC, including the luminal layer where tight junctions are expressed, compared to the loose packing of cells in proximity to the basolateral membrane.  The presence of tight junctions in epithelial cell layers such as MCF7HER2 provides a possible explanation for disparities in both penetration and toxicity between cell lines with variable tumor-packing profiles.  Tight junctions form a semi-permeable barrier in epithelia that regulates the passage of ions and molecules into the paracellular pathway, thus separating the apical and basolateral fluid compartments. In order to cross epithelial sheets and gain access to underlying tissue, drugs must enter either the transcellular or paracellular pathway.  Lipophilic drugs and molecules selectively transported by channels, pumps, and carriers are able to utilize the transcellular route, whereas hydrophilic molecules, such as antibodies, cannot cross biological membranes and are limited to the paracellular pathway  (González-Mariscal, Nava, & Hernández, 2005).   In our current study, where tight junction-containing MCF7HER2 MCC were exposed to trastuzumab in circulating media, immunohistochemical staining of bound antibody revealed little penetration past the first cell layer where tight junctions are present.  Interestingly, at as early as 4hrs of trastuzumab exposure, there were scattered regions of the MCC luminal layer where trastuzumab was seen to penetrate into the tissue beyond the first  69 cell layer.  At increased exposure times, penetration increased and was comparable to the penetration occurring from the basolateral side of the MCC. One explanation for these regions is to identify them as “leaks” in the tight junction barrier, possibly arising from incomplete development or damage to the tight junction network.  Although the loss of tight junction expression in these regions was difficult to confirm in IHC images due to complex ZO-1 branching and the potential occurrence of leakages originating from outside the plane of section, this represents a potential model where the tight junction barrier can be compromised to allow antibody access to the paracellular pathway.  Another well-characterized function of tight junctions is to act as a diffusion barrier to plasma membrane lipids and integral membrane proteins, helping to define apical and basolateral portions of polarized epithelia.  This role is clearly shown in composite IHC images of bound trastuzumab and ZO-1, where trastuzumab binding of HER2 is confined apically in MCC despite constitutive HER2 expression.  Combined, these data suggest the proper functioning of tight junctions at the luminal cell layer of these MCC and the inability of trastuzumab to circumvent the paracellular pathway barrier via intracellular transport.  In contrast to the tight junction-containing luminal cell layer, we observed almost no formation of a tight junction barrier at the membrane-attached, or basolateral, side of the MCF7HER2 MCC.  Trastuzumab was found to penetrate  70 into tissue at rates and distances similar to previous drug penetration studies in MCC composed of non-tight junction expressing cell lines  (Kyle, Huxham, Yeoman, & Minchinton, 2007).  In the MCC model, the thick culture is attached to the 25-30 µm thick, porous Teflon membrane with pore size of 0.4 µm.  Access of drug in the media to the cell culture can therefore be seen as largely unimpeded.  For subsequent trastuzumab penetration, similarities can be drawn to populations of blood vessels in solid breast cancer tumour xenografts where effective distribution of bound trastuzumab around blood vessels has been monitored (Figure 3.1).  Comparing plots of trastuzumab distance distribution over time reveals similar profiles between MCF7HER2 tumour xenografts and the basolateral side of MCC where no ZO-1 is present.  It is important to note, however, that in tumour xenografts distribution was heterogeneous throughout the tumour and blood vessels displayed variable trastuzumab penetration, with some vessels perpetuating little or no extravasation.  Such differences in drug penetration may be attributed to microenvironmental factors such as incomplete or irregular blood vessel development, intermittent blood flow, elevated interstitial fluid pressure, and composition of the extracellular matrix.  Based on the differences in penetration observed between the luminal and basolateral side of the MCF7HER2 MCC, it is plausible that if similar formations of tight junctions occur in epithelial tumours, especially in close association with tumour blood vessels, then tight junctions could represent an additional barrier to populations of cells in the tumour.   71 The loss of tight junctions is believed to be an important event in the stepwise progression of cancers towards the metastatic state.  Multiple studies correlate a loss of various tight junction components with the advancement of breast cancer and other epithelial malignancies (Kominsky, et al., 2003) (Hoover, Liao, & Bryant, 1998) (Chlenski, et al., 1999).  In a tumour metastasis model proposed by Martin and Jiang, cancer cells must overcome tight junctions at three distinct locations – first, to detach from the primary tumour mass where epithelial tight junctions maintain polarity and cohesion of the tissue, thus allowing invasion of the surrounding stroma.  Next, the cancer must intravasate across endothelial tight junctions to gain access to a nearby blood vessel. Finally, the circulating cells must extravasate from the vessel to establish the secondary tumour site  (Martin & Jiang, 2008).  The involvement of tight junctions at any point in this process implies a tight junction barrier that could conceivably hinder drugs administered via the reverse path from the blood to the tumour. Such a situation would constitute opposing roles of tight junctions in tumours, where their loss may increase the metastatic and invasive potential of a cancer but their maintenance may have a negative impact on drug delivery.  In our MCF7HER2 tumour xenografts the barrier function of ZO-1 positive tight junctions was not clearly distinguished.  There was minimal organization and a lack of clustering of ZO-1 positive cells and trastuzumab penetration from vessels was not noticeably influenced by the presence or absence of tight junctions.  Although MCF7HER2 is derived from an adenocarcinoma cell line, the  72 regulation of tight junctions in xenografts may be diminished compared to in situ human breast tissue where ductal and glandular acini, structures whose cells produce the principal tight junctions of the breast, are still present.  Additionally, cell culture xenografts lack the gradient of cancer cells represented at earlier stages of disease progression and differentiation.  Further studies investigating alternate xenograft cell lines or human breast cancer samples may reveal different patterns of tight junction localization with respect to vasculature, thus establishing populations of tumour cells that are inaccessible to antibody-based cancer therapy.   73 Figure 4.1 MCF7HER2 MCC exposed to trastuzumab from the luminal and basolateral sides of the culture MCC were treated with 60 µg/mL trastuzumab for 1-48 hrs.  IHC sections of the MCF7HER2 MCC were stained for HER2 (green) and bound trastuzumab (red). Over time, trastuzumab is seen to penetrate the culture from the bottom (basolateral) side of the culture but is mostly excluded from the top (luminal) side despite equal exposure to trastuzumab during the incubation.  The domes that are formed on the luminal side of the culture are typical for the MCF7 cell line.       74 Figure 4.2 Trastuzumab penetration profile in MCF7HER2 MCC The distance along the x-axis comprises the entire thickness of the MCC.  The majority of staining is localized to the basolateral side of the culture.  This is the side of the tissue disk that, during incubation, was situated adjacent to the porous tissue culture insert membrane.  The side of the culture that was facing free media (represented on the far right of the graph) is here termed the luminal side of the MCC.  Error bars denote mean ± SD.  n = 6.       75 Figure 4.3 Transmission electron micrographs of MCF7HER2 MCC junctional complexes (a) Tight junctions, (c) desmosomes, and (d) interdigitating folds were identified in cellular junctions between cells of the luminal cell layer.  (e) Large intercellular spaces were found separating adjacent cells at the basolateral side of the MCC. (b) Inset highlighting a tight junction kiss point.  (f) Inset emphasizing the size difference of the intercellular gaps of the luminal (b) and basolateral (f) sides.      76 Figure 4.4 Immunohistochemical imaging of ZO-1 positive tight junctions A continuous band of the ZO-1 tight junction marker (red) followed the outer cell layer of tissue background (blue) on the luminal side of the MCC (a).  Deep invaginations extending from the luminal surface were also positive for tight junctions (arrows).  ZO-1 staining was absent from the basolateral layer of cells (b).  Throughout the middle tissue of the MCC, intense ZO-1 staining are present, and may be a result of invaginations cutting through the plane of view.     77 Figure 4.5 Tight junctions on the luminal surface of MCC restrict trastuzumab penetration IHC staining with (b) and without (a) ZO-1 tight junctions (blue) are shown.  The HER2 receptor (green) and bound trastuzumab (red) are also stained in these sections.  The presence of tight junctions limits the penetration of trastuzumab to the outer cell layer of the luminal surface of the MCC.  In some cases, trastuzumab staining is restricted to only one surface of a cell because of tight junction proximity.        78 Figure 4.6 Tight junctions in MCF7HER2 xenografts Immunohistochemical staining of trastuzumab (red), CD31-positive endothelial cells (blue), and ZO-1 tight junctions (green) is shown below.  The heterogeneous distribution of trastuzumab can be seen in (a), where vessels exhibit either high or low (white arrow) extravascular penetration of trastuzumab. (b) Tight junction staining seems sporadic and not localized to any clusters of cells.  (c) The composite image does not reveal any obvious associations of the tight junction staining with either trastuzumab distribution or blood vessel location.    79 5 GENERAL CONCLUSIONS  5.1 Summary and future directions Visualization of trastuzumab-bound HER2 in the MCF7HER2 human breast cancer xenograft model revealed extravascular penetration of trastuzumab up to 150 µm into tumour tissue.  However, heterogeneity of trastuzumab distribution was observed around individual vessels and throughout whole tumour sections, leaving large populations of tumour cells unbound by the antibody treatment. Functionally, trastuzumab treatment decreased expression of the HER2 receptor, delayed tumour growth, and increased vascular permeability of dextrans in a manner consistent with an immune complex-initiated inflammatory response.  Because the precise mechanisms of action caused by trastuzumab treatment are still not fully understood, further research into the trastuzumab- activated inflammatory response should be performed.  Temporarily increased vascular permeability that is accessible to larger biomolecules may also be useful for combination drug therapies.  Additionally, although much research has investigated immune complex-mediated inflammation in normal tissues, very few studies have induced and monitored this effect in tumour xenografts.  The recruitment of activated inflammatory components, such as neutrophils, complement, and platelets, to the tumour microenvironment may be an important consideration for researches attempting to design modified antibody fragments with different immune activating capabilities  (van Egmond, 2008).  80  Treatment of MCF7HER2 tumours with imatinib resulted in increased necrosis, decreased microvessel density, and reduced pericyte expression. Trastuzumab distribution did not improve as previously expected but instead was largely reduced.  Further analysis revealed that blood vessels positive for pericyte coverage possessed greater trastuzumab penetration while the reverse was true for pericyte-negative blood vessels.  Overall, the decreased proportion of pericyte coverage following imatinib treatment appeared responsible for the reduction in trastuzumab levels.  Although the use of imatinib pretreatment to increase the distribution of trastuzumab was not ideal for these tumour xenografts, the identification of a population of cells that are preferentially permeable to trastuzumab, and possibly other drug treatments, warrants further research.  IHC co-staining of other vascular and perivascular markers may further identify these vessels and provide targets for anti-vascular therapy.  Additional work is currently being pursued using magnetic resonance imaging (MRI) to assess tumour perfusion and/or permeability.  These results will be correlated to IHC imaging of multiple vascular markers, including desmin for pericyte coverage.  In vitro MCC of MCF7HER2 cells were treated with trastuzumab from both sides of the culture insert.  Surprisingly, the penetration of trastuzumab into the cultures differed significantly between the membrane bound (basolateral) and  81 luminal surfaces of the tissue disks, the latter demonstrating a barrier function against the antibody.  TEM imaging revealed junctional complexes, including TJ, along the luminal cell layer but absent within the basolateral layer.  IHC staining of ZO-1 confirmed the presence of TJ in a continuous band along the luminal cell layer.  Co-staining of ZO-1 and bound trastuzumab revealed penetration of trastuzumab to be blocked from the paracellular pathway and restricted to the apical portion of luminal cells.  However, in localized instances of an incomplete ZO-1 band, trastuzumab was able to penetrate the luminal layer and access deeper layers of tissue.  Similar analysis was performed in the tumour xenograft model but TJ were found to be inconsistent and haphazard with no distinct patterns capable of a barrier function.   Additional research to detect TJ in either orthotopic mouse breast, alternate xenograft models, or human breast cancer tumour samples may reveal TJ with different localization and pattern formation, especially with respect to blood vessels.  Xenograft models may not be ideal for TJ analysis since their development following inoculation into mice may lack the structural foundation and chemical gradients present in a ductal or lobular carcinoma in situ.  Because the breast microenvironment may contain a higher proportion of normal blood vessels and clearly segregated tissue structures, both expressing ordered TJ patterns compared to their corresponding tumour components, the likelihood of TJ interfering with drug treatment may increase.   82 REFERENCES   Abramsson, A., Berlin, O., Papayan, H., Paulin, D., Shani, M., & Betsholtz, C. (2002). Analysis of mural cell recruitment to tumor vessels. Circulation , 105 (1), 112-7.  Abramsson, A., Lindblom, P., & Betsholtz, C. (2003). Endothelial and nonendothelial sources of PDGF-B regulate pericyte recruitment and influence vascular pattern formation in tumors. The Journal of clinical investigation , 112 (8), 1142-51.  Ambati, J., Canakis, C., Miller, J., Gragoudas, E., Edwards, A., Weissgold, D., et al. (2000). Diffusion of high molecular weight compounds through sclera. Investigative ophthalmology & visual science , 41 (5), 1181-5.  Arthur, J. (2000). The MDCK cell line is made up of populations of cells with diverse resistive and transport properties. Tissue & cell , 32 (5), 446-50.  Baker, J., Lindquist, K., Huxham, L., Kyle, A., Sy, J., & Minchinton, A. (2008). Direct Visualization of Heterogeneous Extravascular Distribution of Trastuzumab in Human Epidermal Growth Factor Receptor Type 2 Overexpressing Xenografts. Clinical Cancer Research , 14 (7), 2171-2179.  Benjamin, L., Golijanin, D., Itin, A., Pode, D., & Keshet, E. (1999). Selective ablation of immature blood vessels in established human tumors follows vascular endothelial growth factor withdrawal. The Journal of clinical investigation , 103 (2), 159-65.  Bergers, G., & Song, S. (2005). The role of pericytes in blood-vessel formation and maintenance. Neuro-oncology , 7 (4), 452-64.  Betsholtz, C., Lindblom, P., & Gerhardt, H. (2005). Role of pericytes in vascular morphogenesis. EXS (94), 115-25.  Booy, E., Johar, D., Maddika, S., Pirzada, H., Sahib, M., Gehrke, I., et al. (2006). Monoclonal and bispecific antibodies as novel therapeutics. Archivum immunologiae et therapiae experimentalis , 54 (2), 85-101.  Brufsky, A. (2010). Trastuzumab-based therapy for patients with HER2-positive breast cancer: from early scientific development to foundation of care. American journal of clinical oncology , 33 (2), 186-95.  Carmeliet, P., & Jain, R. (2000). Angiogenesis in cancer and other diseases. Nature , 407 (6801), 249-57.   83 Cereijido, M., Robbins, E., Dolan, W., Rotunno, C., & Sabatini, D. (1978). Polarized monolayers formed by epithelial cells on a permeable and translucent support. The Journal of cell biology , 77 (3), 853-80.  Chlenski, A., Ketels, K., Engeriser, J., Talamonti, M., Tsao, M., Koutnikova, H., et al. (1999). zo-2 gene alternative promoters in normal and neoplastic human pancreatic duct cells. International journal of cancer Journal international du cancer , 83 (3), 349-58.  Clarijs, R., Otte-Höller, I., Ruiter, D., & de Waal, R. (2002). Presence of a fluid- conducting meshwork in xenografted cutaneous and primary human uveal melanoma. Investigative ophthalmology & visual science , 43 (4), 912-8.  Clynes, R., Towers, T., Presta, L., & Ravetch, J. (2000). Inhibitory Fc receptors modulate in vivo cytoxicity against tumor targets. Nature medicine , 6 (4), 443-6.  Coleman, M., Quaresma, M., Berrino, F., Lutz, J.-M., De Angelis, R., Capocaccia, R., et al. (2008). Cancer survival in five continents: a worldwide population-based study (CONCORD). The lancet oncology , 9 (8), 730-56.  Colombo, M., Corsi, F., Foschi, D., Mazzantini, E., Mazzucchelli, S., Morasso, C., et al. (2010). HER2 targeting as a two-sided strategy for breast cancer diagnosis and treatment: Outlook and recent implications in nanomedical approaches. Pharmacological research : the official journal of the Italian Pharmacological Society , 62 (2), 150-65.  Cooley, S., Burns, L., Repka, T., & Miller, J. (1999). Natural killer cell cytotoxicity of breast cancer targets is enhanced by two distinct mechanisms of antibody- dependent cellular cytotoxicity against LFA-3 and HER2/neu. Experimental hematology , 27 (10), 1533-41.  Cuello, M., Ettenberg, S., Clark, A., Keane, M., Posner, R., Nau, M., et al. (2001). Down-regulation of the erbB-2 receptor by trastuzumab (herceptin) enhances tumor necrosis factor-related apoptosis-inducing ligand-mediated apoptosis in breast and ovarian cancer cell lines that overexpress erbB-2. Cancer research , 61 (12), 4892-900.  Damonte, P., Hodgson, J., Chen, J., Young, L., Cardiff, R., & Borowsky, A. (2008). Mammary carcinoma behavior is programmed in the precancer stem cell. Breast Cancer Research , 10 (3), R50.  Davies, C., Berk, D., Pluen, A., & Jain, R. (2002). Comparison of IgG diffusion and extracellular matrix composition in rhabdomyosarcomas grown in mice versus in vitro as spheroids reveals the role of host stromal cells. British Journal of Cancer , 86 (10), 1639-44.   84 Dejana, E., Lampugnani, M., Martinez-Estrada, O., & Bazzoni, G. (2000). The molecular organization of endothelial junctions and their functional role in vascular morphogenesis and permeability. The International journal of developmental biology , 44 (6), 743-8.  Deli, M. (2009). Potential use of tight junction modulators to reversibly open membranous barriers and improve drug delivery. Biochimica et biophysica acta , 1788 (4), 892-910.  Denekamp, J., & Hobson, B. (1982). Endothelial-cell proliferation in experimental tumours. British Journal of Cancer , 46 (5), 711-20.  dit Faute, M., Laurent, L., Ploton, D., Poupon, M.-F., Jardillier, J.-C., & Bobichon, H. (2002). Distinctive alterations of invasiveness, drug resistance and cell-cell organization in 3D-cultures of MCF-7, a human breast cancer cell line, and its multidrug resistant variant. Clinical & experimental metastasis , 19 (2), 161-8.  Dragsten, P., Blumenthal, R., & Handler, J. (1981). Membrane asymmetry in epithelia: is the tight junction a barrier to diffusion in the plasma membrane? Nature , 294 (5843), 718-22.  Escobar, P., Patrick, R., Rybicki, L., Weng, D., & Crowe, J. (2007). The 2003 revised TNM staging system for breast cancer: results of stage re-classification on survival and future comparisons among stage groups. Annals of surgical oncology , 14 (1), 143-7.  Farquhar, M., & Palade, G. (1963). Junctional complexes in various epithelia. The Journal of cell biology , 17, 375-412.  Flessner, M., Choi, J., Credit, K., Deverkadra, R., & Henderson, K. (2005). Resistance of tumor interstitial pressure to the penetration of intraperitoneally delivered antibodies into metastatic ovarian tumors. Clinical cancer research : an official journal of the American Association for Cancer Research , 11 (8), 3117- 25.  Folberg, R., Hendrix, M., & Maniotis, A. (2000). Vasculogenic mimicry and tumor angiogenesis. The American journal of pathology , 156 (2), 361-81.  Fukumura, D., & Jain, R. (2007). Tumor microenvironment abnormalities: causes, consequences, and strategies to normalize. Journal of Cellular Biochemistry , 101 (4), 937-49.  Gennari, R., Menard, S., Fagnoni, F., Ponchio, L., Scelsi, M., Tagliabue, E., et al. (2004). Pilot study of the mechanism of action of preoperative trastuzumab in patients with primary operable breast tumors overexpressing HER2. Clinical  85 cancer research : an official journal of the American Association for Cancer Research , 10 (17), 5650-5.  Gerhardt, H., & Betsholtz, C. (2003). Endothelial-pericyte interactions in angiogenesis. Cell and tissue research , 314 (1), 15-23.  González-Mariscal, L., Nava, P., & Hernández, S. (2005). Critical role of tight junctions in drug delivery across epithelial and endothelial cell layers. The Journal of membrane biology , 207 (2), 55-68.  Gonzalez-Angulo, A., Hortobágyi, G., & Esteva, F. (2006). Adjuvant therapy with trastuzumab for HER-2/neu-positive breast cancer. The Oncologist , 11 (8), 857- 67.  Grantab, R., Sivananthan, S., & Tannock, I. (2006). The penetration of anticancer drugs through tumor tissue as a function of cellular adhesion and packing density of tumor cells. Cancer research , 66 (2), 1033-9.  Graus-Porta, D., Beerli, R., Daly, J., & Hynes, N. (1997). ErbB-2, the preferred heterodimerization partner of all ErbB receptors, is a mediator of lateral signaling. The EMBO journal , 16 (7), 1647-55.  Guo, P., Hu, B., Gu, W., Xu, L., Wang, D., Huang, H.-J., et al. (2003). Platelet- derived growth factor-B enhances glioma angiogenesis by stimulating vascular endothelial growth factor expression in tumor endothelia and by promoting pericyte recruitment. The American journal of pathology , 162 (4), 1083-93.  Hall, A. (2006). Review of the pericyte during angiogenesis and its role in cancer and diabetic retinopathy. Toxicologic pathology , 34 (6), 763-75.  Hanahan, D., & Weinberg, R. (2000). The hallmarks of cancer. Cell , 100 (1), 57- 70.  Hashizume, H., Baluk, P., Morikawa, S., McLean, J., Thurston, G., Roberge, S., et al. (2000). Openings between defective endothelial cells explain tumor vessel leakiness. The American journal of pathology , 156 (4), 1363-80.  Hassid, Y., Eyal, E., Margalit, R., Furman-Haran, E., & Degani, H. (2008). Non- invasive imaging of barriers to drug delivery in tumors. Microvascular research , 76 (2), 94-103.  Hasumi, Y., Kłosowska-Wardega, A., Furuhashi, M., Ostman, A., Heldin, C.-H., &  Hellberg, C. (2007). Identification of a subset of pericytes that respond to combination therapy targeting PDGF and VEGF signaling. International journal of cancer Journal international du cancer , 121 (12), 2606-14.  86  Helmlinger, G., Netti, P., Lichtenbeld, H., Melder, R., & Jain, R. (1997). Solid stress inhibits the growth of multicellular tumor spheroids. Nature biotechnology , 15 (8), 778-83.  Hennessy, B., Smith, D., Ram, P., Lu, Y., & Mills, G. (2005). Exploiting the PI3K/AKT pathway for cancer drug discovery. Nature reviews Drug discovery , 4 (12), 988-1004.  Hiscox, S., Davies, E., & Barrett-Lee, P. (2009). Aromatase inhibitors in breast cancer. Maturitas , 63 (4), 275-9.  Hoover, K., Liao, S., & Bryant, P. (1998). Loss of the tight junction MAGUK ZO-1 in breast cancer: relationship to glandular differentiation and loss of heterozygosity. The American journal of pathology , 153 (6), 1767-73.  Hormigo, A., Gutin, P., & Rafii, S. (2007). Tracking normalization of brain tumor vasculature by magnetic imaging and proangiogenic biomarkers. Cancer cell , 11 (1), 6-8.  Huang, W., Newman, B., Millikan, R., Conway, K., Hulka, B., Schell, M., et al. (2000). Risk of breast cancer according to the status of HER-2/neu oncogene amplification. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology , 9 (1), 65-71.  Hughes, R., Kurth, M., McGilligan, V., McGlynn, H., & Rowland, I. (2008). Effect of colonic bacterial metabolites on Caco-2 cell paracellular permeability in vitro. Nutrition and cancer , 60 (2), 259-66.  Hynes, N., & Lane, H. (2005). ERBB receptors and cancer: the complexity of targeted inhibitors. Nature Reviews Cancer , 5 (5), 341-54.  Inai, T., Mancuso, M., Hashizume, H., Baffert, F., Haskell, A., Baluk, P., et al. (2004). Inhibition of vascular endothelial growth factor (VEGF) signaling in cancer causes loss of endothelial fenestrations, regression of tumor vessels, and appearance of basement membrane ghosts. The American journal of pathology , 165 (1), 35-52.  International Agency for Research on Cancer. (2008). World Cancer Report. Retrieved from http://www.iarc.fr/en/publications/pdfs-online/wcr/  Izumi, Y., Xu, L., di Tomaso, E., Fukumura, D., & Jain, R. (2002). Tumour biology: herceptin acts as an anti-angiogenic cocktail. Nature , 416 (6878), 279- 80.   87 Jain, R. (2005). Normalization of tumor vasculature: an emerging concept in antiangiogenic therapy. Science (New York, NY) , 307 (5706), 58-62.  Kominsky, S., Argani, P., Korz, D., Evron, E., Raman, V., Garrett, E., et al. (2003). Loss of the tight junction protein claudin-7 correlates with histological grade in both ductal carcinoma in situ and invasive ductal carcinoma of the breast. Oncogene , 22 (13), 2021-2033.  Kyle, A., Huxham, L., Chiam, A., Sim, D., & Minchinton, A. (2004). Direct assessment of drug penetration into tissue using a novel application of three- dimensional cell culture. Cancer research , 64 (17), 6304-9.  Kyle, A., Huxham, L., Yeoman, D., & Minchinton, A. (2007). Limited tissue penetration of taxanes: a mechanism for resistance in solid tumors. Clinical cancer research : an official journal of the American Association for Cancer Research , 13 (9), 2804-10.  Lane, H., Motoyama, A., Beuvink, I., & Hynes, N. (2001). Modulation of p27/Cdk2 complex formation through 4D5-mediated inhibition of HER2 receptor signaling. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO , 12 Suppl 1, S21-2.  Laughner, E., Taghavi, P., Chiles, K., Mahon, P., & Semenza, G. (2001). HER2 (neu) signaling increases the rate of hypoxia-inducible factor 1alpha (HIF-1alpha) synthesis: novel mechanism for HIF-1-mediated vascular endothelial growth factor expression. Molecular and cellular biology , 21 (12), 3995-4004.  Leu, A., Berk, D., Lymboussaki, A., Alitalo, K., & Jain, R. (2000). Absence of functional lymphatics within a murine sarcoma: a molecular and functional evaluation. Cancer research , 60 (16), 4324-7.  Lin, A., Rugo,, H. (2007). The role of trastuzumab in early stage breast cancer: current data and treatment recommendations. Current treatment options in oncology , 8 (1), 47-60.  Lister, K., James, W., & Hickey, M. (2007). Immune Complexes Mediate Rapid Alterations in Microvascular Permeability: Roles for Neutrophils, Complement, and Platelets. Microcirculation , 14 (7), 709-722.  Love, S.M. (2005). Dr. Susan Love’s Breast Book. (4th ed.). Cambridge, MA: Da Capo Press.  Maass, N., Alkasi, O., Bauer, M., Jonat, W., Souchon, R., & Meinhold-Heerlein, I. (2009). Actual management of ductal carcinoma in situ of the breast. Archives of gynecology and obstetrics , 280 (5), 699-705.   88 Maniotis, A., Folberg, R., Hess, A., Seftor, E., Gardner, L., Pe'er, J., et al. (1999). Vascular channel formation by human melanoma cells in vivo and in vitro: vasculogenic mimicry. The American journal of pathology , 155 (3), 739-52.  Marasanapalle, V., Li, X., Polin, L., & Jasti, B. (2006). Novel in vitro model barriers for evaluation of the permeability of antitumor compounds, thioxanthones. Investigational New Drugs , 24 (2), 111-6.  Martin, T., & Jiang, W. (2008, Nov 14). Loss of tight junction barrier function and its role in cancer metastasis. Biochimica et biophysica acta .  Martin, T., Mansel, R., & Jiang, W. (2002). Antagonistic effect of NK4 on HGF/SF induced changes in the transendothelial resistance (TER) and paracellular permeability of human vascular endothelial cells. Journal of cellular physiology , 192 (3), 268-75.  Martin, T., Watkins, G., Mansel, R., & Jiang, W. (2004). Hepatocyte growth factor disrupts tight junctions in human breast cancer cells. Cell biology international , 28 (5), 361-71.  Martin, T., Watkins, G., Mansel, R., & Jiang, W. (2004). Loss of tight junction plaque molecules in breast cancer tissues is associated with a poor prognosis in patients with breast cancer. European Journal of Cancer , 40 (18), 2717-2725.  Massarweh, S., & Schiff, R. (2006). Resistance to endocrine therapy in breast cancer: exploiting estrogen receptor/growth factor signaling crosstalk. Endocrine- related cancer , 13 Suppl 1, S15-24.  Matter, K., & Balda, M. (2003). Functional analysis of tight junctions. Methods (San Diego, Calif) , 30 (3), 228-34.  McCann, M., Gill, C., O' Brien, G., Rao, J., McRoberts, W., Hughes, P., et al. (2007). Anti-cancer properties of phenolics from apple waste on colon carcinogenesis in vitro. Food and chemical toxicology : an international journal published for the British Industrial Biological Research Association , 45 (7), 1224- 30.  McDonald, D., & Baluk, P. (2002). Significance of blood vessel leakiness in cancer. Cancer research , 62 (18), 5381-5.  McDonald, D., & Foss, A. (2000). Endothelial cells of tumor vessels: abnormal but not absent. Cancer metastasis reviews , 19 (1-2), 109-20.  Minchinton, A., & Tannock, I. (2006). Drug penetration in solid tumours. Nature Reviews Cancer , 6 (8), 583-592.   89 Mori, M., Sawada, N., Kokai, Y., & Satoh, M. (1999). Role of tight junctions in the occurrence of cancer invasion and metastasis. Medical electron microscopy : official journal of the Clinical Electron Microscopy Society of Japan , 32 (4), 193- 198.  Morikawa, S., Baluk, P., Kaidoh, T., Haskell, A., Jain, R., & McDonald, D. (2002). Abnormalities in pericytes on blood vessels and endothelial sprouts in tumors. The American journal of pathology , 160 (3), 985-1000.  Nahta, R., Yu, D., Hung, M.-C., Hortobagyi, G., & Esteva, F. (2006). Mechanisms of Disease: understanding resistance to HER2-targeted therapy in human breast cancer. Nature Clinical Practice Oncology , 3 (5), 269-280.  Netti, P., Berk, D., Swartz, M., Grodzinsky, A., & Jain, R. (2000). Role of extracellular matrix assembly in interstitial transport in solid tumors. Cancer research , 60 (9), 2497-503.  Nicolson, G. (1988). Cancer metastasis: tumor cell and host organ properties important in metastasis to specific secondary sites. Biochimica et biophysica acta , 948 (2), 175-224.  Ostman, A. (2004). PDGF receptors-mediators of autocrine tumor growth and regulators of tumor vasculature and stroma. Cytokine & Growth Factor Reviews , 15 (4), 275-286.  Parkin, D., Bray, F., Ferlay, J., & Pisani, P. (2005). Global cancer statistics, 2002. CA: a cancer journal for clinicians , 55 (2), 74-108.  Pervez, S., Kirkland, S., Epenetos, A., Mooi, W., Evans, D., & Krausz, T. (1989). Effect of polarity and differentiation on antibody localization in multicellular tumour spheroid and xenograft models and its potential importance for in vivo immunotargeting. International journal of cancer Journal international du cancer , 44 (5), 940-7.  Petit, A., Rak, J., Hung, M., Rockwell, P., Goldstein, N., Fendly, B., et al. (1997). Neutralizing antibodies against epidermal growth factor and ErbB-2/neu receptor tyrosine kinases down-regulate vascular endothelial growth factor production by tumor cells in vitro and in vivo: angiogenic implications for signal transduction therapy of solid tumors. The American journal of pathology , 151 (6), 1523-30.  Pietras, K., & Ostman, A. (2010). Hallmarks of cancer: interactions with the tumor stroma. Experimental cell research , 316 (8), 1324-31.  Pritchard, K. (2005). Endocrinology and hormone therapy in breast cancer: endocrine therapy in premenopausal women. Breast Cancer Research , 7 (2), 70-6.  90  Rae, J., Creighton, C., Meck, J., Haddad, B., & Johnson, M. (2007). MDA-MB- 435 cells are derived from M14 melanoma cells--a loss for breast cancer, but a boon for melanoma research. Breast Cancer Research and Treatment , 104 (1), 13-9.  Ren, J., Hamada, J., Takeichi, N., Fujikawa, S., & Kobayashi, H. (1990). Ultrastructural differences in junctional intercellular communication between highly and weakly metastatic clones derived from rat mammary carcinoma. Cancer research , 50 (2), 358-62.  Ribatti, D., & Vacca, A. (2008). The role of microenvironment in tumor angiogenesis. Genes & nutrition , 3 (1), 29-34.  Rocha, A., Azevedo, I., & Soares, R. (2008). Progesterone sensitizes breast cancer MCF7 cells to imatinib inhibitory effects. Journal of Cellular Biochemistry , 103 (2), 607-14.  Romond, E., Perez, E., Bryant, J., Suman, V., Geyer, C., Davidson, N., et al. (2005). Trastuzumab plus adjuvant chemotherapy for operable HER2-positive breast cancer. The New England journal of medicine , 353 (16), 1673-84.  Ross, J., & Fletcher, J. (1998). The HER-2/neu Oncogene in Breast Cancer: Prognostic Factor, Predictive Factor, and Target for Therapy. The Oncologist , 3 (4), 237-252.  Rossi, S., Marciello, M., Sandri, G., Bonferoni, M., Ferrari, F., & Caramella, C. (2008). Chitosan ascorbate: a chitosan salt with improved penetration enhancement properties. Pharmaceutical development and technology , 13 (6), 513-21.  Rothen-Rutishauser, B., Braun, A., Günthert, M., & Wunderli-Allenspach, H. (2000). Formation of multilayers in the caco-2 cell culture model: a confocal laser scanning microscopy study. Pharmaceutical research , 17 (4), 460-5.  Rucker, H., Wynder, H., & Thomas, W. (2000). Cellular mechanisms of CNS pericytes. Brain research bulletin , 51 (5), 363-9.  Russo, J., Bradley, R., McGrath, C., & Russo, I. (1977). Scanning and transmission electron microscopy study of a human breast carcinoma cell line (MCF-7) cultured in collagen-coated cellulose sponge. Cancer research , 37 (7 Pt 1), 2004-14.  Sandoval, K., & Witt, K. (2008). Blood-brain barrier tight junction permeability and ischemic stroke. Neurobiology of disease , 32 (2), 200-19.   91 Sgroi, D. (2010). Preinvasive breast cancer. Annual review of pathology , 5, 193- 221.  Slamon, D., Clark, G., Wong, S., Levin, W., Ullrich, A., & McGuire, W. (1987). Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science (New York, NY) , 235 (4785), 177-82.  Soler, A., Miller, R., Laughlin, K., Carp, N., Klurfeld, D., & Mullin, J. (1999). Increased tight junctional permeability is associated with the development of colon cancer. Carcinogenesis , 20 (8), 1425-31.  Tait, C., Waterworth, A., Loncaster, J., Horgan, K., & Dodwell, D. (2005). The oligometastatic state in breast cancer: hypothesis or reality. Breast (Edinburgh, Scotland) , 14 (2), 87-93.  Tallquist, M., & Kazlauskas, A. (2004). PDGF signaling in cells and mice. Cytokine & Growth Factor Reviews , 15 (4), 205-13.  Thomlinson, R., & Gray, L. (1955). The histological structure of some human lung cancers and the possible implications for radiotherapy. British Journal of Cancer , 9 (4), 539-49.  Thurber, G., Schmidt, M., & Wittrup, K. (2008). Antibody tumor penetration: Transport opposed by systemic and antigen-mediated clearance. Advanced drug delivery reviews , 60 (12), 1421-34.  Tobioka, H., Sawada, N., Zhong, Y., & Mori, M. (1996). Enhanced paracellular barrier function of rat mesothelial cells partially protects against cancer cell penetration. British Journal of Cancer , 74 (3), 439-45.  Toffoli, S., & Michiels, C. (2008). Intermittent hypoxia is a key regulator of cancer cell and endothelial cell interplay in tumours. The FEBS journal , 275 (12), 2991- 3002.  Tong, R., Boucher, Y., Kozin, S., Winkler, F., Hicklin, D., & Jain, R. (2004). Vascular normalization by vascular endothelial growth factor receptor 2 blockade induces a pressure gradient across the vasculature and improves drug penetration in tumors. Cancer research , 64 (11), 3731-6.  Tsukita, S., Yamazaki, Y., Katsuno, T., & Tamura, A. (2008). Tight junction- based epithelial microenvironment and cell proliferation. Oncogene , 27 (55), 6930-8. van Egmond, M. (2008). Neutrophils in antibody-based immunotherapy of cancer. Expert opinion on biological therapy , 8 (1), 83-94.   92 Vlahovic, G., Ponce, A., Rabbani, Z., Salahuddin, F., Zgonjanin, L., Spasojevic, I., et al. (2007). Treatment with imatinib improves drug delivery and efficacy in NSCLC xenografts. British Journal of Cancer , 97 (6), 735-40.  Vlahovic, G., Rabbani, Z., Herndon, J., Dewhirst, M., & Vujaskovic, Z. (2006). Treatment with Imatinib in NSCLC is associated with decrease of phosphorylated PDGFR-β and VEGF expression, decrease in interstitial fluid pressure and improvement of oxygenation. British Journal of Cancer , 95 (8), 1013-1019.  Warburton, C., Dragowska, W., Gelmon, K., Chia, S., Yan, H., Masin, D., et al. (2004). Treatment of HER-2/neu overexpressing breast cancer xenograft models with trastuzumab (Herceptin) and gefitinib (ZD1839): drug combination effects on tumor growth, HER-2/neu and epidermal growth factor receptor expression, and viable hypoxic cell fraction. Clinical cancer research : an official journal of the American Association for Cancer Research , 10 (7), 2512-24.  Willett, C., Boucher, Y., di Tomaso, E., Duda, D., Munn, L., Tong, R., et al. (2004). Direct evidence that the VEGF-specific antibody bevacizumab has antivascular effects in human rectal cancer. Nature medicine , 10 (2), 145-7.  Wilson, C., Cajulis, E., Green, J., Olsen, T., Chung, Y., Damore, M., et al. (2005). HER-2 overexpression differentially alters transforming growth factor-beta responses in luminal versus mesenchymal human breast cancer cells. Breast Cancer Research , 7 (6), R1058-79.  Winkler, F., Kozin, S., Tong, R., Chae, S.-S., Booth, M., Garkavtsev, I., et al. (2004). Kinetics of vascular normalization by VEGFR2 blockade governs brain tumor response to radiation: role of oxygenation, angiopoietin-1, and matrix metalloproteinases. Cancer cell , 6 (6), 553-63.  Wolff, A., Hammond, M., Schwartz, J., Hagerty, K., Allred, D., Cote, R., et al. (2007). American Society of Clinical Oncology/College of American Pathologists guideline recommendations for human epidermal growth factor receptor 2 testing in breast cancer. Archives of pathology & laboratory medicine , 131 (1), 18-43.  World Health Organization. (2009). Cancer [Fact Sheet]. Retrieved from http://www.who.int/mediacentre/factsheets/fs297/en/index.html  Yakes, F., Chinratanalab, W., Ritter, C., King, W., Seelig, S., & Arteaga, C. (2002). Herceptin-induced inhibition of phosphatidylinositol-3 kinase and Akt Is required for antibody-mediated effects on p27, cyclin D1, and antitumor action. Cancer research , 62 (14), 4132-41.  Yarden, Y., & Sliwkowski, M. (2001). Untangling the ErbB signalling network. Nature reviews Molecular cell biology , 2 (2), 127-37.   93 APPENDIX   Animal Care Certificate   


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items